The current study sets out to explore test anxiety in adolescent students. The effect of sociodemographic variables
on test anxiety was controlled for and the relationship between test anxiety and other psychological constructs,
such as self-criticism, social anxiety, acceptance and mindfulness, was examined. In addition, the predictive
effect/power of these variables was analyzed and a comparative study between high and low test anxiety
adolescents was conducted. Participants in this study were 449 high school students, 211 boys and 238 girls,
with a mean age of 16.28 years. These participants completed a battery of self-report questionnaires composed
by the Portuguese versions of Test Anxiety Inventory (TAI), Child Acceptance and Mindfulness Measure
(CAMM), Forms of Self-Criticizing/Attacking and Self-Reassuring Scale (FSCRS), and the Social Anxiety and
Avoidance Scale for Adolescents (SAASA). Results showed that gender, self-criticism and competencies for
acceptance and mindfulness had a significant and an independent contribution on the prediction of test anxiety.
The comparative study revealed that adolescents with high test anxiety score significantly higher in negative
forms of self-criticism, social anxiety and lower in self-reassurance, acceptance and mindfulness, when compared
to those with low test anxiety. Despite its exploratory nature, the current study adds to the existing knowledge
on the influence of psychological processes, such as self-criticism and acceptance, on test anxiety. These findings
might constitute a relevant contribution to psychological intervention with adolescents.
Keywords: test anxiety, adolescence, self-criticism, acceptance and mindfulness.
Este estudio se propone explorar la ansiedad ante los exámenes en adolescentes. El efecto de las variables
sociodemográficas fue controlado y se examinó la relación entre esta forma de ansiedad y la autocrítica, la
ansiedad social, la aceptación y la conciencia plena. Además, se analizó el poder predictivo de estas variables
y se realizó un estudio comparativo entre adolescentes con ansiedad ante los exámenes alta y baja. Participaron
449 alumnos de Educación Secundaria, 211 varones y 238 mujeres, con una edad media de 16.28 años. Los
instrumentos de medida utilizados han sido: Test Anxiety Inventory (TAI), Child Acceptance and Mindfulness
Measure (CAMM), Forms of Self-Criticizing/Attacking and Self-Reassuring Scale (FSCRS) y la Escala de
Ansiedad y Evitación de Situaciones Sociales para Adolescentes (EAESSA). Los resultados mostraron que el
género, la autocrítica y las competencias para la aceptación y atención plena tuvieron un efecto significativo y
una contribución independiente sobre la predicción de la ansiedad ante los exámenes. El estudio comparativo
reveló que los adolescentes con alta ansiedad ante los exámenes puntúan significativamente más alto en las
formas negativas de autocrítica y de ansiedad social, y muestran niveles más bajos de autoconfianza, aceptación
y conciencia plena, comparados con aquellos adolescentes con bajos niveles de ansiedad ante los exámenes.
A pesar de su naturaleza exploratoria, este estudio completa los conocimientos existentes sobre la influencia
de procesos psicológicos, como la autocrítica y la aceptación, en la ansiedad ante los exámenes. Estos hallazgos
podrían así constituir una contribución relevante para la intervención psicológica con adolescentes.
Palabras clave: ansiedad ante los exámenes, adolescencia, autocrítica, aceptación, conciencia plena.
Text Anxiety in Adolescents: The Role of Self-Criticism
and Acceptance and Mindfulness Skills
Marina Cunha1and Maria Jacinta Paiva2
1Instituto Universitário de Coimbra (Portugal)
2Universidade de Coimbra (Portugal)
The Spanish Journal of Psychology
2012, Vol. 15, No. 2, 533-543
Copyright 2012 by The Spanish Journal of Psychology
Correspondence concerning this article should be addressed to Marina Cunha. Instituto Superior Miguel Torga, Coimbra; CINEICC
- Centro de Investigação do Núcleo de Estudos e Intervenções Cognitivo-Comportamentais da Universidade de Coimbra. Largo Cruz
de Celas, n.º1, 3000-132 Coimbra (Portugal). Phone: +351-239488030. Fax: +351-239488031. E-mail: firstname.lastname@example.org
In Portugal, in university teaching context, exam anxiety
appears as one of the problems that motivate a larger
number of students to ask for support in psycho-pedagogic
counseling facilities (Pereira, Masson, Ataíde, & Melo,
2004; Melo, Pinto-Gouveia, & Pereira, 2006). Despite the
absence of empiric studies on high school teenagers, it is
consensual that the perception of this phenomenon is very
frequent. It is in this school stage that students decide the
application to university, and school grade competition
assumes a major role, generating greater anxiety. Hence,
we have deliberately chosen high school adolescent
population as the object of this study.
Nowadays, scientific literature widely accepts anxiety as
inherent to human condition, and as an adaptation tool, a
response to real or symbolic danger. Exam anxiety related to
anxiety in school or academic evaluation situations is a further
example of the alarm process involved in anticipation or
confrontation with a situation one perceives as “dangerous”,
as very difficult and exceeding one’s skills or resources to
deal with it (Beck, Emery, & Greenberg, 1985; Zeidner, 1998).
Evaluation anxiety, present in all individuals, in a smaller or
larger scale, may have an adaptation function if it helps to
concentrate resources and abilities to produce more and better
in an effective manner, or, on the contrary, it may assume a
non-adaptive function, if it decreases efficacy and/or prevents
the completion of tasks and challenges desired by the subject
(or necessary to the achievement of goals). According to
theoreticians of evaluation anxiety, its main manifestations
involve concerns related to a possible negative performance
and respective consequences (referring to evaluation others
might do about the self, and or to consequences for the future)
and high levels of emotionality (expressed by an increased
physiologic activation) (Liebert & Morris, 1967).
In a theoretical point of view, two definitive contributions
have emerged from the wide range of studies undertaken
between the 60s and the 80s. On the one hand, Spielberger
and Vagg (1995) developed the distinction operated by
Cattell and Scheier between trace anxiety and state anxiety.
State anxiety would be transitory, occurring in occasional
situations, such as a particularly difficult exam, or for which
the student was not properly prepared; while trace anxiety
would constitute a psychological predisposition to react with
a certain level of anxiety (high or low) to an indiscriminate
number of situations of evaluation or threat.
On the other hand, Liebert and Morris (1967), in an
innovative study, decomposed exams’ anxiety in two
fundamental factors: cognitive (worry) and affective
(emotionality). Since then, the multidimensional conception
of exams’ anxiety (Zeidner, 1998) assumed an important
heuristic value in studies guidelines (Sarason, 1984, 1986,
1988). In fact, the bi-dimensional structure identified by
Liebert and Morris (1967) was adopted by Spielberger in
Test Anxiety Inventory (Spielberger, et al., 1980). Later
investigations (e.g., Deffenbacher & Hazaleus, 1985)
indicate that the cognitive component (namely, worry) exerts
a preponderant action in the deterioration of academic
In general, anxious students are extremely vulnerable
to contextual signs of evaluation situations; before, during
and after tests/exams, the student feels insecure, worried
and doubtful (Zeidner, 1998). Intrusive thoughts (e.g.,
anticipation of failure) interfere with the attention processes
required by the task and, consequently, jeopardize
performance (Sarason, 1984, 1988). As a matter of fact,
studies conducted in this area of research have shown the
negative effect that excessive anxiety has in performance
situations, mostly on the academic level (Cruz, 1989;
McDonald, 2001; Prins & Hanewalds, 1997; Seipp, 1991)
when, in most cases, individuals possess the necessary
knowledge to effectively perform the tasks.
On a clinical perspective, anxiety towards tests, being
primarily a preoccupation with a negative evaluation, can
be framed within the diagnosis of Social Anxiety Disorder,
as it is described by DSM (American Psychiatric Association,
2002). Worrying about the scrutiny of others is a central
aspect of social anxiety disorder, whether it is centered in
school assessments, whether it is generalized to other social
situations of interaction and performance (Cunha, Pinto-
Gouveia, & Soares, 2007), causing a meaningful negative
impact in the several areas of the individual’s life (school,
work, family, society). In adolescence, the assessment of
excessive or irrational fear from social situations may
become especially critical, if we take into account the various
development tasks unique of this age group that favor a
greater vulnerability to experiences of social anxiety (Cunha,
Pinto-Gouveia, & Salvador, 2008).
In his first studies, Sarason (1984, 1986) had already
drawn the attention to the role performed by criticism in
the development and maintenance of anxiety to tests/exams.
In this line of thought, recent studies about self-criticism
(Gilbert, 2005, 2009; Gilbert, Clarke, Hempel, Miles, &
Irons 2004; Irons, Gilbert, Baldwin, Baccus, & Palmer,
2006) may supply useful clues concerning the role of this
concept in psychopathology.
Gilbert (2000) suggests that human beings evolved
towards being more and more responsive to social signs
by others. Nevertheless, during the evolution process,
humans may have become more self-conscious and able to
think about themselves also in relational terms. In this
manner, the nature of our internal dialogue became relational
and our self-evaluations are adaptations of assessment
processes originally designed for social interactions, so that
the signs generated internally have the same physiological
and emotional impact than social signs emitted externally.
From this perspective, if we possess the ability of
representing different social roles, we also have different
parts of the Self when distinct cognitive-emotional and
motivational patterns emerge in the mind. These internal
parts of the Self may have different voices and can compete,
agree or disagree, support or attack mutually.
CUNHA AND PAIVA
Hence, self-criticism appears as a psychological process
which translates scrutiny and censorship of personal
behaviors, thoughts and emotions (Whelton & Greenberg
2005). Through two distinct comparisons, although not
contrary, the self-critical individual devalues and looks
down at his/her qualities (Irons et al., 2006). However,
comparison with others (compared self-criticism), or
comparison with certain patterns of personal demands
(internalized self-criticism) settle the subjective perception
of a gap between real and ideal.
In the case of adolescents with tests’ anxiety, internal
attacks often take place when they sense they have not
matched a certain pattern (e.g., “I don’t understand any of
this! Why can’t I solve this? I can’t go far …!”). The
adolescent actually believes that this self-criticism process
will press them to do better or to correct the mistakes. Melo
(2006), in her study conducted in university context,
evidenced the importance of self-criticism on the development
of exam anxiety.
Another contribution that seemed interesting to explore
for the intervention on tests’ anxiety originates from
acceptance based therapeutic approaches and on the concept
of mindfulness (Farmer & Chapman, 2008). In childhood
and adolescence, this type of intervention has gained ground,
showing to be especially effective in anxiety disorders
(Greco, Blackledge, Coyne, & Ehrenreich, 2005). Acceptance
skills and self-consciousness (mindfulness) consist of a
persistent attitude of being in the present moment; a constant
exercise of attention to the present, identifying thoughts,
feelings and body sensations without judging them.
Therefore, this kind of psychological process inverts the
usual way of functioning and replaces the descending
information processing by the ascending, outlining heuristic
filters: memories, beliefs and cognitive distortions (Brown,
Ryan, Creswell, & Niemiec, 2008). Scrutiny is compassionate
and diametrically opposite to the inflexibility that
characterizes self-criticism. The intervention based on this
model lies upon the acceptance of internal experience and
does not aim at the reduction of symptoms, but the
commitment with important values and the pursuit in the
sense of living according to them (Eifert, & Forsyth, 2005).
The concept of psychological acceptance refers to
openness and availability to experience private events (e.g.,
body sensations, thoughts, feelings, memories) as they are,
without struggle or defense (Hayes, Strosahl, & Wilson,
1999), assuming them as a process opposite to experiential
avoidance. On the other hand, experiential avoidance is an
omnipresent process, learnt in early years and reinforced
by socio-verbal community throughout life, responsible for
psychological inflexibility, which leads to the exaggeration
of human suffering. A range of researches has shown a
positive association between psychopathologic symptoms
and efforts made to avoid emotions, thoughts, memories
and other private events (Hayes et al., 1999; Eifert, &
In the case of adolescents with exam anxiety, due to
the frequency of intrusive and dysfunctional thoughts, on
one side, and the high emotional levels, on the other, we
believe it is important to develop concentration skills with
purpose and without critical judgment (mindfulness skills),
as well as promote acceptance of internal experiences,
imperfections, less successful performances, and difficult
The present investigation intends to analyze the
contribution of these new constructs, such as self-criticism,
acceptance and mindfulness skills, in order to understand
and intervene on test anxiety. From this point of view, we
analyzed the relationship between these concepts, their
predictive capacity and verify in what extent individuals
with high anxiety to tests distinguish from individuals with
low anxiety, in the way they criticize themselves, self-
evaluate and observe and accept their inner experiences.
The sample includes 449 high school teenagers, 211
boys (47%) and 238 girls (53%) distributed by the three
school grades (10th, 11thand 12thgrade), with a mean age
of 16.8 (DP = 1.17).
Subjects are equitably distributed by school years and
by age groups, with exception of the 18-21-year-old group,
which presents, as expected, a minor number of individuals.
The socio-economic level of teenagers was determined
by the average of parents’ categories, with a predominant
high socioeconomic level (62.6%) in this sample.
There were no differences between boys and girls in
what concerns their distribution by school years [χ2(2) =
4.48; p = .107], by age groups [χ2(3) = 6.26; p = .10)], and
by socioeconomic level [χ2(2) = 1.23; p = .541].
The following instruments were used: a) socio-
demographic questionnaire to determine age, gender, school
grade and socio-economic level; b) TAI (Spilberger et al.,
1980) for exam anxiety assessment; c) FSCRS (Gilbert et
al., 2004) to evaluate forms of self-criticism; SAASA to
assess social anxiety in adolescents (Cunha, Pinto-Gouveia,
Salvador, & Alegre 2004) and d) CAMM (Greco, Smith, &
Baer, 2008) to measure acceptance and mindfulness in
The Test Anxiety Inventory (TAI- Test Anxiety Inventory,
Spilberger, et al., 1980; Portuguese version: Ponciano
Loureiro, Pereira, & Speilberger, 2005) was conceived by
Spielberger to evaluate individual differences of test anxiety
in an academic context, taking it as a personality feature
related with this specific situation.
TEST ANXIETY IN ADOLESCENTS
The inventory consists of 20 items that aim to analyze
the frequency with which anxiety symptoms are experienced,
before, during and prior tests and exams. These items are
distributed by two factors concerning worrying (W – worry)
and emotionality (E – emotionality). Worrying involves the
expression of fear of performance (e.g., to think about failure
consequences), and Emotionality refers to physiological and
affective reactions to test situation stress (e.g., feeling
apprehension, worry, agitation or the heart beating fast).
Items are answered on a 4-point Likert scale which varies
from “Hardly ever” to “Almost always”. The global score
may vary between 20 and 80 points and the anxiety level
will be proportionate to the number of points.
The Test Anxiety Inventory was translated and adapted
by Ponciano and colleagues for university Portuguese
students, revealing good psychometric properties, namely
internal consistency (α = .93) and temporal reliability (r =
.69; p < .01), and the same factorial structure of the original
version (Ponciano et al., 2005).
For the present investigation, Ponciano’s version has
undergone slight changes to better adapt to the studied
sample. Some concepts have been reformulated, once they
did not relate to high school teaching; items language was
adapted in order to make them clearer for this school level
students. In our sample, this inventory revealed a good
internal consistency, showing a Cronbach alpha value of .92
(Worry subscale: α = .84; Emotionality subscale: α = .87).
Self-criticism Forms and Self-Reassuring Scale (FSCRS
- Forms of Self-Criticizing and Self-Reassuring Scale, Gilbert
et al., 2004; Portuguese version: Castilho & Pinto-Gouveia,
2005) consists of a set of 42 items to evaluate how people
criticize and reassure themselves “when things go wrong”.
Participants answer to a range of situations, in a 5-point Likert
scale (0 – I’m not like that to 4 – I’m exactly like that). This
measure is composed by three factors: the Inadequate Self,
assessing the feeling of inadequacy of the self when facing
failure and recess (e.g. “I get easily disappointed with
myself”); the Reassuring Self , which evaluates the capacity
of the self to reassure, to look for comfort, to calm down and
to be self-compassionate (e.g., “I can remind myself of my
positive characteristics”); e and the Hated Self, assessing the
feeling of self-loath/hate and a destructive response when
facing failures, characterized by an aggressive persecution to
hurt him/herself (e.g., “I get so angry I want to hurt myself”).
The higher the score in each factor, the more frequent is the
corresponding self-criticism form.
For this investigation, and taking into consideration the
mean age group (15-18 years-old) of our sample, we were
cautious to reformulate some items, in order to make them
more understandable for our participants.
In the current study, internal consistency values obtained
in this scale were: .87 for the Inadequate Self sub-scale
(original version, in English: .90), .82 for the Reassuring
Self sub-scale (.86 in the original version, English) and .76
for the Hated Self sub-scale (.86 original version, English).
The Social Anxiety and Avoidance Scale for Adolescents
(SAASA; Cunha et al., 2004) aims to assess the discomfort
and avoidance level in a wide range of social situations
representative of the most frequent social fears in adolescents
(Cunha et al., 2004; Cunha, Pinto-Gouveia, & Salvador,
2007, Cunha et al., 2008).
SAASAis a self-report instrument composed by 34 items,
describing social situations. All items are rated on a five-point
Likert scale: anxiety subscale (1 – no anxiety; 2 – few anxiety;
3 – a little anxiety; 4 – a lot of anxiety; 5 – great anxiety)
and the avoidance subscale (1 – never; 2 – sometimes; 3 –
often; 4 – most of the time; 5 – almost always).
The total score, for each subscale, may range from 34
and 170 points. Higher scores indicate higher levels of anxiety
and avoidance. SASSA showed good psychometric properties
and a six factors structure was found in several studies: (1)
Interaction in new social situations; (2) Interaction with the
opposite sex; (3) Performance in formal social situations;
(4) Assertive interaction; (5) Being observed by others; (6)
Eat and drink in public (Cunha et al., 2004, Cunha et al.,
2007, Cunha et al., 2008)
In this study, SAASA also presented an excellent internal
consistency with Cronbach’s alpha of .94 and .92 for the
anxiety and avoidance sub-scale, respectively.
Children’s Acceptance and Mindfulness Measure
(CAMM; Children’s Acceptance and Mindfulness
Measure, Greco et al., 2008; Portuguese version: Cunha,
Pinto-Gouveia, & Paiva, 2010) was initially developed
by Greco and Baer (2006) with the goal of evaluating
concerning their ability to observe inner experience, to
consciously accept their own internal experiences. It is
a 25-item self-report instrument answered in a 5-point
Likert scale: never – 0; rarely – 1; sometimes – 2; almost
– 3; always – 4. Based on studies about item content
and exploratory factor analysis, this scale was recently
reformulated to a version consisting only of 10 items
(Greco et al., 2008).
In the 10-item single-factor scale, scores vary from 0 to
40 points and the higher the score the higher is the
acceptance of the individual and the consequent mindfulness
Previous studies suggest that CAMM presents good
internal consistency (α = .80) and good concurrent validity.
In our sample this instrument (10 items version) revealed
an acceptable internal consistency, with Cronbach’s alpha
values of .70.
After collecting the proper authorizations for the
consecution of the study, the assessment battery was
distributed to a pre-test group of 15 students from other
schools. This pre-test confirmed the clarity of the items and
allowed to realize the filling mean time.
CUNHA AND PAIVA
The questionnaires were answered in group, in classroom
context. Participation was voluntary and anonymity and
confidentiality assured. The total filling in of questionnaires
took in average 40 minutes.
Statistical data analysis
In what concerns data analysis, we have used the
statistic data analysis program SPSS (15.0 version).
Statistical procedures were chosen based on Pestana
and Gageiro (2003) recommendations. Parametric tests were
used because our sample size justified them. A significant
level of .05 was considered (Howell, 2006).
Depending on comparison groups (gender, age, school
grade and test/exam anxiety) t-tests or analysis of variance
(ANOVA) were carried out for mean comparisons regarding
X2were calculated to compare frequencies between the
different groups. We used Pearson coefficient to establish
relations between the studied variables, and the hierarchic
regression analysis to investigate the group of variables
most likely to predict exam anxiety.
Influence of the socio-demographic variables on
When comparing test anxiety values depending on the
socio-demographic variables (Table 1), using t-tests and
ANOVA, we verified that there are significant gender
differences (t = -4.25; p = .001), and socioeconomic level
differences [F(2, 442) = 8.46; p = .001]. Girls, comparatively
to boys, reveal higher anxiety scores (M = 45.43; DP =
11.19) and middle socioeconomic level adolescents present
greater anxiety (M = 46.40; DP = 10.83), presenting
significantly higher scores compared to high socioeconomic
level adolescents (p = .001). There was no relevant effect
of age groups [F(3, 441) = .47; p = .706] and of school
grade on test anxiety [F(2, 442) = 1.36; p = .259].
Given the importance of gender in result differentiation,
we also investigated its effect on the remaining studied
variables. On table 2 we can see the obtained mean values
in the self-report instruments according to gender. There are
significant statistical differences between boys and girls
concerning test/exam anxiety (TAI - total and emotionality
factor, respectively: t = -4.25; p = < .05; t = -5.03; p < .05),
self-criticism Inadequate Self dimension (t = -3.98; p < .05),
TEST ANXIETY IN ADOLESCENTS
Effects of gender, age, school grade, and socio-economic
level variables on test anxiety measured by TAI
M SD F/tp
15 years old
Means and standard deviations of studied variables according to gender
M SDM SDtp
TAI – Emotionality
FSCRS - Hated self
FSCRS - Reassuring self
FSCRS - Inadequate self
SAASA – Total Avoidance
?ote. TAI: Test Anxiety Inventory; FSCRS: Forms of Self-Criticizing and Self-Reassuring Scale; SASSA: Social Anxiety and Avoidance
Scale for Adolescents; CAMM: Children’s Acceptance and Mindfulness Measure
correspondently: t = -5.08; p < .05 and t = -4.36; p < .05),
and, last, acceptance and mindfulness (t = 3.93; p < .05).
Except for acceptance and mindfulness dimension, in which
boys (M = 23.70; DP = 5.63) obtain higher scores than girls
(M = 21.70; DP = 4.78), the opposite pattern occurs in the
In summary, in our sample, girls present higher
test/exam anxiety, social anxiety, more inadequacy feelings
towards failure and less acceptance and mindfulness
features, when compared to boys.
anxiety (anxietyand avoidancesubscales,
Relationship between test/exam anxiety and social
anxiety, self-criticism, and acceptance and
As can be observed in Table 3, test/exam anxiety
correlates positively and significantly with self-criticism
factors of the Inadequate Self (r = .57; p < .01) and the
Hated Self (r = .38; p < .01), as well as with anxiety and
social avoidance (respectively, r = .30; p < .01 e r = .25;
p < .01). It also reveals a negative significant correlation
with acceptance and mindfulness dimension (r = -.36; p <
.01) and the Reassuring Self (r = -.21, p < .01).
In other words, we can state that in adolescents the
greater the test/exam anxiety, the greater the feeling of
inadequacy of the self towards failure, the greater the feeling
of self-disgust and destructive answer when facing failures,
and the greater the social anxiety. By the contrary,
acceptance and mindfulness skills decrease, as well as the
capacity of reassuring and being compassionate to oneself.
Comparison study between test/exam anxiety
Aiming at understanding in which variables students
with higher test/exam anxiety levels differentiated, two
groups were formed based on median TAI values.
Adolescents with scores inferior to the median constituted
the low test/exam anxiety group (LTA), and students whose
scores were superior to the median formed the high test/exam
anxiety group (HTA). Since significant differences were
CUNHA AND PAIVA
Correlations between studied variables
Total TAIHated Self
FSCRS - Hated self
FSCRS - Reassuring self
FSCRS - Inadequate self
SAASA – Total Avoidance
?ote. TAI: Test Anxiety Inventory; FSCRS: Forms of Self-Criticizing and Self-Reassuring Scale; SASSA: Social Anxiety and Avoidance
Scale for Adolescents; CAMM: Children’s Acceptance and Mindfulness Measure
** Significant correlation from .01
Means and standard deviations of psychological measures according to comparison groups
Low TA (n = 213)
High TA (n = 214)
FSCRS - Hated Self
FSCRS - Reassuring Self
FSCRS ¬ Inadequate Self
SAASA –Total Anxiety
SAASA –Total Avoidance
?ote. Low TA: Low Test Anxiety; High TA: High Test Anxiety; TAI: Test Anxiety Inventory; FSCRS: Forms of Self-Criticizing and
Self-Reassuring Scale; SASSA: Social Anxiety and Avoidance Scale for Adolescents; CAMM: Children’s Acceptance and Mindfulness
observed in TAI results according to gender, this cutoff point
based on median values was found separately for boys and
girls (boys: Mdn = 39; girls: Mdn = 45). Subjects with
median TAI scores were excluded from the groups. The low
test/exam anxiety Group (LTA) represented 213 students
(98 boys and 115 girls) and the high anxiety Group (HTA)
was constituted by 214 teenagers (103 boys and 111 girls).
When the two test/exam anxiety groups are compared
(Table 4), one observes a significant difference between
them in all studied variables, showing that students with
high test/exam anxiety equally have higher levels of
negative self-criticism (inadequate self and Hated self),
higher social anxiety scores (anxiety and avoidance), as
well as lower acceptance and mindfulness scores and a
minor reassuring capacity (reassuring self).
Predicting test/exam anxiety variables
To investigate the predicting value of self-criticism and
accepting/mindfulness skills, concerning test/exam anxiety
(total TAI), a Multiple Hierarchic Regression Analysis was
A Multiple Hierarchic Regression function was calculated,
being composed by two blocks, corresponding to the nature
of independent variables considered as relevant for the
prediction of the dependent variable. This approach allows
investigating whether the prediction can be improved by a
variable, or group of variables, according to the effect of
Based on the described goal, we inserted the gender
variable in the first block, since we found, in previous analysis,
gender differences in several psychological measures. In the
second block we inserted variables relative to the perception
of the individuals about the way they criticize themselves
(Hated self, inadequate and reassuring), and also about their
acceptance of internal experiences and mindfulness skills.
Knowing that models for the formulation of multiple
regression linear equations implicate the numeric character
of dependent and independent variables, some adjusting
procedures were necessary for gender. According to
Pedhazur (1997), the problem about categorical variables
(nominal or ordinal) may be solved with indicator variables
As can be seen in Table 5, the first block of regression
function explains 4,3% of the total variance and the second
block contributed with 29,5%. The function as a whole
explains 33,7% (R2= .337) of the total variance [F(5, 412)
= 41.92; p = .001] of test anxiety.
Going to standardized regression coefficients (Beta
values) that translate the predictive value of the different
variables considered in the model (Table 6), gender variable
reveals to be, in the first block, a predictive factor of
test/exam anxiety. Based on the fact Beta values are positive,
(β = .206; p < .001), we can state that being female (the
indicator category) seems to be significantly associated to
higher levels of test/exam anxiety.
When the two blocks are introduced, the higher Beta value
is for the inadequate self (β = .437; p = .001), gender variable
(β = .110; p = .009) and, finally, acceptance/mindfulness
variable (β = -.098; p = .036). Hence, we can conclude that
higher test/exam anxiety levels measured by TAI are associated
to higher levels of negative self-criticism (inadequate self),
female gender and low acceptance and levels.
Given the high frequency of youngsters with test/exam
anxiety and the disturbing character this phenomenon may
assume, causing school, personal and social problems, we
believe it is important to further study this subject in order
to promote useful and effective intervention strategies to
face these challenges.
Previous Portuguese studies, in university student samples,
have shown test/exam anxiety as the main reason for
specialized psychological support (Pereira, et al., 2004; Melo
et al., 2006). We do not have empirical data, in high school
TEST ANXIETY IN ADOLESCENTS
Summary of hierarchic multiple regression for test/exam anxiety measured by TAI
Regression coefficient on test/exam anxiety measured by TAI
1. Gender 4.727 .206 4.304 .001
Acceptance and Mindfulness
–0.184 –.012 –0.258
–0.213 –.098 –2.108
context, on test/exam anxiety, in Portugal, although it is
consensual that it is a common and disturbing problem. In
an international plan, excessive anxiety in evaluation
situations, namely tests, exams and oral presentations, is one
the most prevailing conditions in children and adolescents
(Beidel, & Turner, 1988; McDonald, 2001). Notwithstanding
this information relevance, research, and therapeutic
intervention even more, do not seem to accompany its high
prevalence. In this line of thought, and under the light of
recent therapeutic practices, like mindfulness and acceptance
therapy and compassion therapy (Farmer & Chapman, 2008),
our study aimed to analyze the role of these psychological
processes about test/exam anxiety in high school adolescents.
Self-report instruments used in the investigation were
subjected to slight adaptations for adolescents and their
psychometric characteristics were studied in this specific
sample, showing adequate qualities of precision and validity.
When analyzing the influence of socio-demographic
variables, results show that girls manifest higher test/exam
anxiety than boys, which is consistent with the literature
on the subject. Among children, girls usually denote greater
anxiety (cf. Beidel & Turner, 2006) and, according to DSM-
IV-TR (APA 2002), epidemiological studies indicate that
social anxiety is higher among women, although clinical
samples nullify or reverse this effect. The meta-analysis
study conducted by Seipp and colleagues (Seipp, 1991
Seipp & Schwarzer, 1996) with data obtained from the TAI
in 14 different countries points to this clear trend of women
with higher test/exam anxiety levels than men.
This pattern is most prominent in the emotionality factor
scores than in the worry dimension results. Eventually, this
cross-cultural trend is rooted in social stereotypes (Silvern
& Katz 1986) which consider shyness acceptable and even
desirable in women as in men it is unacceptable, and in
socialization practices that encourage women to express their
emotions and men to suppress them, including anxiety. Also,
in Portugal, Melo’s study (2006) on college students found
similar results in TAI, according to gender. We find, therefore,
the effect of gender, which marks the studies on anxiety,
particularly among adolescents (e.g., Cunha et al., 2004).
Concerning the role of the socioeconomic level, young
people of average socioeconomic status exhibited greater
test/exam anxiety, compared to the ones with a high
socioeconomic level. Although adolescents with a low
socioeconomic status presented higher levels of test/exam
anxiety than the ones with a high level, these two groups
did not differ significantly. The interpretation of these results
requires some caution, since the low socioeconomic level
in our sample was under-represented. In general, literature
gives higher anxiety values for students coming from
disadvantaged socio-economic classes (Zeidner 1998, Melo
2006). According to Zeidner (1998), these students enhance
their perception of danger and insecurity to face challenges,
developing, therefore, high levels of anxiety, when they
verify they have fewer resources to cope with school
demands than their colleagues. It is important to remember
that adolescents in our study were attending high school,
grade in which there are national standardized exams and
whose results put intense pressure on teachers, parents and
students. The results obtained in high school education are
dependent on the means of application to higher education,
this process being increasingly competitive, selective and
anxiety-inducing. The resource to tutorials or additional
paid classes to better prepare students, the possibility of
access to private higher education institutions, in case they
cannot have high grades for the desired course, are often
resources lower classes do not have.
Age and school grade have not shown any relevant
effect on test/exam anxiety, suggesting the subject
homogeneity of our sample in developing terms.
The study on the relationship between test/exam anxiety
and the remaining psychological variables revealed test/exam
anxiety is positively connected to self-criticism forms, such
as inadequate self and hated self, to social anxiety and
avoidance. It also showed a significant correlation, although
negative, with acceptance/mindfulness dimension and with
the reassuring Self. In other words, in adolescents, the greater
the test/exam anxiety, the greater the feeling of inadequacy
of the self when facing failures, the greater the feeling of
self-disgust and of a destructive answer towards failure,
greater social anxiety and smaller capacity of acceptance/
mindfulness, as well as smaller ability to reassure and be
compassionate to oneself. These data comply with literature
data, namely, the ones referring to positive association
between social anxiety and test anxiety, highlighting the
common evaluation anxiety base (Beidel, & Turner, 1988;
Cunha et al., 2007, Zeidner, 1998). On the other hand, in
what concerns self-criticism, Sarason (1984) and Melo’s
studies (2006) had brought to evidence the same type of
notwithstanding the different assessment instruments used
in the various studies. Finally, concerning the acceptance/
mindfulness process, there are no studies so far using these
instruments, given the very recent development of these
measures for adolescents. Nevertheless, other studies (Greco
et al., 2005; Greco, Smith et al., 2008; Greco, Lambert, &
Baer, 2008) have demonstrated that subjects with low
acceptance and mindfulness skills present more anxiety
problems, confirming the present investigation results.
In order to know in which variables students with low
and high test/exam anxiety differed, two groups were created
based on TAI results and gender. Comparison of these two
groups revealed that the group of students with high
test/exam anxiety showed significantly higher levels of
negative self-criticism (inadequate self and hated self),
higher levels of social anxiety (anxiety and avoidance) and
lowest values of acceptance and mindfulness and the
capacity for self-reassurance (reassuring self).
In a different perspective, we also tried to analyze the
effect of the whole action of test/exam anxiety variables,
CUNHA AND PAIVA
using a hierarchical multiple regression analysis. According
to the results, gender, self-criticism and acceptance/
mindfulness skills are the variables that best predict test
anxiety. Thus, we can state that being female, having high
levels of negative self-criticism (inadequate self) and low
acceptance/mindfulness skills are associated with greater test
anxiety tests, evaluated by TAI, in adolescents.
These results highlight that the way individual self-criticize,
and their acceptance/mindfulness skills, translated by
awareness in the present moment and acceptance of internal
experiences (e.g., thoughts, emotions, bodily sensations) may
have a role in the development and maintenance of test/exam
anxiety. Recriminatory criticism helps to maintain and
exacerbate insecurity, negative automatic thoughts, irrelevant
and intrusive that affect test performance. In this way, they
hinder the adoption of effective study strategies, as well as
an appropriate test approach.
Limitations and future studies
This study, although exploratory, intends to provide
clues and investigate the relationships between the variables
under consideration in order to contribute to a better
understanding of this frequent phenomenon responsible for
the suffering of many adolescents.
A possible implication of our results involves the
intervention with these students, in addition to the usual
strategies, specifically contemplating critical self-evaluation
aspects, promoting forms of reassurance acceptance. Thus,
it may be important to help the adolescent to identify the
internal process of self-criticism, deal with it, and learn to
accept failures as part of human nature and forgiving it
inside a new affective structure. It is worth investing in the
development of intervention strategies to improve the
response given by support facilities to students.
We are aware of the weaknesses of our study for its
methodology and that a thorough study would involve the
observation of other contextual and institutional variables
and the use of more sophisticated methodologies.
Notwithstanding the limitations of this study inherent
to the cross-sectional methodology used, which does not
allow the establishment of causal relationships between
variables, and data collection exclusively by self-report
instruments, we believe to have contributed to a better
understanding of this reality among Portuguese high school
American Psychiatric Association (2002). Manual de diagnóstico
e estatística das perturbações mentais [Diagnostic and
statistical manual of mental disorders] (4thEd.). Lisboa,
Portugal: Climepsi Editores.
Beck, A. T., Emery, G., & Greenberg, R. L. (1985). Anxiety
disorders and phobias: A cognitive perspective. New York,
NY: Basic Books.
Beidel, D. C., & Turner, S. M. (1988). Comorbitidity of test anxiety
and other anxiety disorders in children. Journal of Abnormal
Child Psychology, 16, 275–287. http://dx.doi.org/10.1007/
Beidel, D. C, & Turner, S. M. (2006). Shy children, phobic adults:
?ature and treatment of social anxiety disorders. Washington,
Bodas, J., & Ollendick, T. H. (2005). Test anxiety: A cross-cultural
perspective. Clinical Child and Family Psychology Review,
8, 65–88. http://dx.doi.org/10.1007/s10567-005-2342-x
Brown, K. W., Ryan, R. M., Creswell, J. D., & Niemiec, C. P.
(2008). Beyond me: Mindful responses to social threat. In H.
A. Wayment & J. J. Bauer (Eds.), Transcending self-interest:
Psychological explorations of the quiet ego (pp.75-84).
Washington, DC: American Psychological Association.
Castilho, P., & Pinto-Gouveia, J. (2005). A versão Portuguesa da
escala Forms of Self-Criticizing and Self-Reassuring Scale
[The Portuguese version of the scale of Self-Criticizing Forms
and Self-Reassuring Scale]. Coimbra, Portugal: Faculdade de
Psicologia e Ciências da Educação da Universidade de
Cruz, J. F. (1989). Incidência, desenvolvimento e efeitos da
ansiedade aos exames nos testes e exames escolares [Incidence,
development and effects of test anxiety in school exams].
Revista Portuguesa de Educação, 2, 111–130.
Cunha, M., Pinto-Gouveia, J., Salvador, M. C., & Alegre, S.
(2004). Medos sociais na adolescência: A Escala de Ansiedade
e Evitamento de Situações Sociais para Adolescentes
(EAESSA) [Socials fears in adolescence: The Social Anxiety
and Avoidance Scale for Adolescents (SAASA)]. Psychologica,
Cunha, M., Pinto-Gouveia, J., & Salvador, M. C. (2007). A Escala
de Ansiedade e Evitamento de Situações Sociais para
Adolescentes (EAESSA) [The Social Anxiety and Avoidance
Scale for Adolescents (SAASA)]. In M. R. Simões, C.
Machado, M. M. Gonçalves, & L. S. Almeida (Eds.), Avaliação
psicológica: Instrumentos validados para a população
Portuguesa (pp. 57-76). Coimbra, Portugal: Quarteto.
Cunha, M., Pinto-Gouveia, J., & Soares, I. (2007). Natureza,
frequência e consequências dos medos sociais na adolescência:
dados na população portuguesa [Nature, frequency and
consequences of social fears in adolescence: data from
Portuguese population]. Psychologica, 44, 207–236.
Cunha, M., Pinto-Gouveia, J., & Salvador, M. C. (2008). Social
fears in adolescence: The Social Anxiety and Avoidance Scale
for Adolescents (SAASA). European Psychologist, 13, 197–
Cunha, M., Pinto-Gouveia, J., & Paiva, M. J. (2010, October).
Mindfulness skills in Portuguese adolescents: Psychometric
properties of the Children’s Acceptance and Mindfulness
Measure (CAMM). Paper presented at the meeting of the 40th
Annual Congress of EABCT, Milan, Italy.
TEST ANXIETY IN ADOLESCENTS
Deffenbacher, J. L., & Hazaleus, S. L. (1985). Cognitive,
emotional, and physiological components of test anxiety.
Cognitive Therapy and Research, 9, 169–180. http://dx.doi.
Farmer, R. F., & Chapman, A. L. (2008). Behavioral interventions
in cognitive behavior therapy: Practical guidance for putting
theory into action (pp. 251-278). Washington, DC: APA Books.
Eifert, G. H., & Forsyth, J. P. (2005). Acceptance & commitment
therapy for anxiety disorders: A practitioner’s treatment guide
to using mindfulness, acceptance, and values-based behavior
change strategies. Oakland, CA: New Harbinger Publications.
Gilbert, P. (2000). The relationship of shame, social anxiety and
depression: The role of the evaluation of social rank. Clinical
Psychology and Psychotherapy, 7(3), 174–189. http://dx.doi.
Gilbert, P. (2005). Compassion: Conceptualisations, research and
use in psychotherapy. London - New York: Routledge.
Gilbert, P. (2009). The compassionate mind: Coping with the
challenges of living. London, England: Constable & Robinson.
Gilbert, P., Clarke, M., Hempel, S., Miles, J., & Irons, C. (2004).
Criticizing and reassuring oneself: An exploration of forms,
styles and reasons in female students. British Journal of
Clinical Psychology, 43, 31–50. http://dx.doi.org/10.1348/0144
Greco, L. A., & Baer, R. A. (2006). Child acceptance and
mindfulness measure (CAMM). University of Missouri,
Greco, L. A., Blackledge, J. T., Coyne, L. W., & Ehrenreich, J.
(2005). Integrating acceptance and mindfulness into treatments
for child and adolescent anxiety disorders: Acceptance and
commitment therapy as an example. In S. M. Orsillo & L.
Roemer (Eds.), Acceptance and mindfulness-based approaches
to anxiety: Conceptualization and treatment. (pp. 301-322).
New York, NY: Springer.
Greco, L. A., Smith, G., & Baer, R. A. (2008). Assessing
mindfulness in children and adolescents: Development and
validation of the Children´s Acceptance and Mindfulness
Measure (CAMM). (Unpublished manuscript). University of
Missouri-St. Louis, MO.
Greco, L. A., Lambert, W., & Baer, R. A. (2008). Psychological
inflexibility in childhood and adolescence: Development and
evaluation of the Avoidance and Fusion Questionnaire for
Youth. Psychological Assessment, 20, 93–102. http://dx.doi.org/
Hayes, S. C., Strosahl, K., & Wilson, K. G. (1999). Acceptance
and commitment therapy: An experiential approach to behavior
change. New York, NY: Guilford Press.
Howell, D. (2006). Statistical methods for psychology (6thEd.).
Belmont, CA: Thomson Wadsworth.
Irons, C., Gilbert, P., Baldwin, M. W., Baccus, J. R., & Palmer,
M. (2006). Parental recall, attachment relating and self-
attacking/self-reassurance: Their relationship with depression.
British Journal of Clinical Psychology, 45, 297–308.
Liebert, R. M., & Morris, L. W. (1967). Cognitive and emotional
components of test anxiety: A distinction and some initial
data. Psychological Reports, 20, 975–978. http://dx.doi.org/10.
McDonald, A. S. (2001). The prevalence and effects of test anxiety
in school children. Educacional Psychology, 21, 89–101.
Melo, A. (2006). Ansiedade aos exames em contexto universitário
[Test Anxiety in a university context]. (Unpublished Master’s
thesis). Faculdade de Psicologia e de Ciências da Educação
da Universidade de Coimbra, Portugal.
Melo, A., Pinto-Gouveia, J., & Pereira, A. (2006). Ansiedade aos
exames: Impacto na saúde mental dos estudantes universitários
[Test Anxiety: impact on mental health of college students].
In I. Leal, J. Pais-Ribeiro, & S. Jesus (Eds.), Actas do 6º
Congresso ?acional de Psicologia da Saúde (pp. 123). Lisboa
Portugal: ISPA edições.
Pedhazur, E. J. (1997). Multiple regression in behavioral research:
Explanation and prediction (3rdEd.). Forth Worth, TX:
Harcourt Brace College Publishers.
Pereira, A., Masson, A., Ataíde, R., & Melo, A. (2004). Stresse,
ansiedade e distúrbios emocionais em estudantes universitários
[Stress, anxiety and emotional disorders in college students]. In
J. Pais-Ribeiro & I. Leal (Eds.), Actas do 5º Congresso de Psico-
logia da Saúde (pp.119-125). Lisboa, Portugal: ISPA edições.
Pestana, M. J., & Gageiro, J. N. (2003). Análise de dados para
ciências sociais. A complementariedadede do SPSS [Analysis
of data for social sciences. The complementarity of the SPSS].
Lisboa, Portugal: Sílabo.
Ponciano, E., Loureiro, L., Pereira, A., & Spielberger, C. (2005).
Características psicométricas e estrutura factorial do TAI de
Spielberger em estudantes universitários [Psychometric
characteristics and factor structure of the TAI of Spielberger
in a college students]. In A. S. Pereira & E. D. Motta (Eds.),
Acção social e aconselhamento no ensino superior,
investigação e intervenção – Actas do Congresso ?acional.
Coimbra, Portugal: SASUC.
Prins, P. J., & Hanewald, G. J. (1997). Self-statements of text-
anxious children: Thought-listing and questionnaire approaches.
Journal of Consulting and Clinical Psychology, 65, 440–447.
Sarason, I. G. (1984). Stress, anxiety and cognitive interference:
Reactions to tests. Journal of Personality and Social Psycho-
logy, 46, 929–938. http://dx.doi.org/10.1037//0022-35188.8.131.529
Sarason, I. G. (1986). Test anxiety, worry, and cognitive
interference. In R. Schwarzer (Ed.), Self-related cognitions in
anxiety and motivation, (pp. 19-34). Hillsdale, NJ: L. Erlbaum
Sarason, I. G. (1988). Anxiety, self-preoccupation and attention.
Anxiety Research, 1, 3–7. http://dx.doi.org/10.1080/10615808
Seipp, B., & Schwarzer, C. (1996). Cross-cultural anxiety research:
A review. In C. Schwarzer & M. Zeidner (Eds.), Stress, anxiety,
and coping in academic settings (pp. 13-68). Tubingen,
CUNHA AND PAIVA
Seipp, B. (1991). Anxiety and academic performance: A meta- Download full-text
analysis of findings. Anxiety Research, 4, 27–41. http://dx.doi.
Silvern, L. E., & Katz, P. A. (1986). Gender roles and adjustment
in elementary-school children: A multidimensional approach.
Sex Roles, 14, 181–202. http://dx.doi.org/10.1007/BF00288248
Spielberger, C. D., Gonzalez, H. P., Taylor, C. J., Algaze, B., Ross,
G. R., & Westberry, L. G. (1980). Test Anxiety Inventory. Palo
Alto, CA: Consulting Psychologists Press.
Spielberger, C. D. & Vagg, P. R. (1995). Test anxiety: A transactional
process. In C. D. Spielberg & P. R. Vagg (Eds.), Test anxiety:
Theory, research and application (pp. 3-13). Washington, DC:
Taylor & Francis.
Whelton, W. J., & Greenberg, L. S. (2005). Emotion in self-
criticism. Personality and Individual Differences, 38, 1583–
Zeidner, M. (1998). Test anxiety: The state of the art. New York,
NY: Plenum Press.
Received March 9, 2010
Revision received May 16, 2011
Accepted May 21, 2011
TEST ANXIETY IN ADOLESCENTS