Conscientiousness Is the Most Powerful Noncognitive Predictor of School Achievement in Adolescents

Abstract

Much research has demonstrated that intelligence and conscientiousness have a high impact on individual school achievement. To figure out if other noncognitive traits have incremental validity over intelligence and conscientiousness, we conducted a study on 498 eighthgrade students from general secondary schools in Austria. Hierarchical regressions for three criteria (GPA, science, and languages) were performed, including intelligence, the Big Five, self-discipline, grit, self-efficacy, intrinsic-extrinsic motivation, and test anxiety. Intelligence and conscientiousness alone accounted for approximately 40% in the variance of school achievement. For languages and GPA, no other personality and motivational predictors could explain additional variance; in science subjects, only self-discipline added incremental variance. We conclude that - in addition to intelligence as powerful cognitive predictor - conscientiousness is the crucial noncognitive predictor for school achievement and should be focused on when supporting students in improving their performance.

Full text

Volume 37 / Number 1 / 2016
Journal of
Individual
Dierences
Editor-in-Chief
André Beacuducel
Associate Editors
Philip J. Corr
Sam Gosling
Jürgen Hennig
Philipp Y. Herzberg
Aljoscha Neubauer
Thomas Rammsayer
Willibald Ruch
Stefan Schmukle
Astrid Schütz
Andrzej Sekowski
Jutta Stahl
Martin Vocarek
Journal of Individual Dierences
Volume 37 / Number 1 / 2016
jid_55-0-41-3_58-60_positiv.indd 3 09.12.2015 11:14:30

Contents
Original Articles Gymnasts and Orienteers Display Better Mental Rotation Performance
Than Nonathletes
1
Mirko Schmidt, Fabienne Egger, Mario Kieliger, Benjamin Rubeli,
and Julia Schu
¨ler
Conscientiousness Is the Most Powerful Noncognitive Predictor
of School Achievement in Adolescents
8
Barbara Dumfart and Aljoscha C. Neubauer
Perception of Emotional Expressions in Adults: The Role
of Temperament and Mood
16
Chit Yuen Yi, Matthew W. E. Murry, and Amy L. Gentzler
The Relationships Between the Dark Triad, the Moral Judgment Level,
and the Students’ Disciplinary Choice: Self-Selection, Indoctrination,
or Both?
24
Annika Krick, Stephanie Tresp, Mirijam Vatter, Antonia Ludwig,
Michael Wihlenda, and Martin Rettenberger
Social Support, Emotional Intelligence, and Posttraumatic Stress
Disorder Symptoms: A Mediation Analysis
31
Nicole L. Hofman, Austin M. Hahn, Christine K. Tirabassi,
and Raluca M. Gaher
Rank-Order Consistency and Profile Stability of Self- and Informant-
Reports of Personal Values in Comparison to Personality Traits
40
Henrik Dobewall and Toivo Aavik
Personality of Clown Doctors: An Exploratory Study 49
Alberto Dionigi
Validation and Revision of a German Version of the Balanced Measure
of Psychological Needs Scale
56
Andreas B. Neubauer and Andreas Voss
Ó2016 Hogrefe Publishing Journal of Individual Differences 2016; Vol. 37(1)

Journal of
Individual
Differences
Your article has appeared in a journal published by Hogrefe Publishing.
This e-offprint is provided exclusively for the personal use of the authors. It may not be
posted on a personal or institutional website or to an institutional or disciplinary repository.
If you wish to post the article to your personal or institutional website or to archive it
in an institutional or disciplinary repository, please use either a pre-print or a post-print of
your manuscript in accordance with the publication release for your article and our
‘‘Online Rights for Journal Articles’’ (www.hogrefe.com/journals).

Original Article
Conscientiousness Is the Most
Powerful Noncognitive Predictor of
School Achievement in Adolescents
Barbara Dumfart
1
and Aljoscha C. Neubauer
2
1
Academy of Lower Austria, Sankt Pölten, Austria,
2
Department of Psychology, University of Graz, Austria
Abstract. Much research has demonstrated that intelligence and conscientiousness have a high impact on individual school achievement.
To figure out if other noncognitive traits have incremental validity over intelligence and conscientiousness, we conducted a study on 498 eighth-
grade students from general secondary schools in Austria. Hierarchical regressions for three criteria (GPA, science, and languages) were
performed, including intelligence, the Big Five, self-discipline, grit, self-efficacy, intrinsic-extrinsic motivation, and test anxiety. Intelligence
and conscientiousness alone accounted for approximately 40% in the variance of school achievement. For languages and GPA, no other
personality and motivational predictors could explain additional variance; in science subjects, only self-discipline added incremental variance.
We conclude that – in addition to intelligence as powerful cognitive predictor – conscientiousness is the crucial noncognitive predictor for
school achievement and should be focused on when supporting students in improving their performance.
Keywords: school achievement, intelligence, Big Five, motivation, anxiety
Individual academic achievement in adolescents is deter-
mined by various factors. Across different studies, one of
the most important predictors is intelligence. Depending
on the measure used, the average correlation between intel-
ligence and school grades is about 0.5 (cf. Gustafsson &
Undheim, 1996; Laidra, Pullmann, & Allik, 2007). Some
studies show even higher relationships; in a 5-year prospec-
tive longitudinal study of about 70.000 English children
(Deary, Strand, Smith, & Fernandes, 2007), a correlation
of 0.81 between the gfactor of intelligence and a latent trait
of educational achievement was observed. Especially in
lower grades and nonselective comprehensive schools,
intelligence explains the largest amount of variance in
school achievement. At higher levels of education, such
as the tertiary educational system (college, university),
intelligence is not as important anymore in the prediction
of achievement (Chamorro-Premuzic & Furnham, 2005;
Jensen, 1980). This might go back to the restriction of range
because intelligence has already served as selection
criterion for the admission to the higher educational track
(Boekaerts, 1995).
Intelligence generally seems to be more important for
achievement in science subjects than in languages. In sci-
ence, logical analysis plays a great role, whereas in arts,
especially in languages, traits like social confidence are
essential (Furnham, Rinaldelli-Tabaton, & Chamorro-
Premuzic, 2011).
Even if intelligence is doubtlessly an important factor,
there are many other variables which should be considered
in the prediction of school achievement. Especially when it
comes to educational consulting, it is important to focus not
only on the rather unmalleable trait intelligence, but also on
intrapersonal strengths like personality and motivational
variables. Noncognitive variables might be easier to train
and more sensitive to intervention (Stankov, Lee, Luo, &
Hogan, 2012). One of the central noncognitive variables
to predict school achievement is conscientiousness. In a
meta-analysis it has been shown that conscientiousness is
the most consistent and stable personality predictor for aca-
demic achievement (Poropat, 2009). It combines various
traits which are crucial for successful learning: for example,
self-discipline, ambition, persistence, diligence, and dutiful-
ness. The narrow traits of conscientiousness can predict
academic achievement better than the broad trait (Paunonen
& Ashton, 2001). Duckworth and Seligman (2005) found
out that self-discipline accounted for more than twice as
much variance as intelligence in school achievement and
learning behavior of eighth-grade students. However, this
result could be partly due to the fact that the study was con-
ducted in a selective school and the consequential range
restriction of intelligence.
A conceptually related trait, which has lately been
researched, is grit, defined as the ‘‘perseverance and pas-
sion for long-term-goals’’ (Duckworth, Peterson, Matthews,
& Kelly, 2007, p. 1). Grit integrates aspects of achievement
striving, self-control, and consistency of interests and
encourages the realization of existing talents in an individual.
Duckworth et al. (2007) conducted several studies in
Journal of Individual Differences 2016; Vol. 37(1):8–15
DOI: 10.1027/1614-0001/a000182
2016 Hogrefe Publishing
Author’s personal copy (e-offprint)

high-achieving persons and found out that grit was related to
educational attainment and career stability.
Openness and related traits, such as Typical Intellectual
Engagement or Intellectual Curiosity,havealsoturnedout
to be important for academic achievement. Students who
enjoy spending time with cognitively demanding tasks are
more likely to perform well in school (Goff & Ackerman,
1992; Poropat, 2009; von Stumm, Hell, & Chamorro-Prem-
uzic, 2011). On the other hand, extraversion turned out as a
trait whose impact on academic achievement changes with
age. While extraversion might be beneficial for school
performance in earlier years of formal education, in higher
levels of education it is negatively correlated with grades.
Thiscouldbetracedbacktothechanging–fromsocial
to formal – character of school across different levels of
education (O’Connor & Paunonen, 2007). A similar pattern
is found for agreeableness: in primary school, it seems to
have relatively high impact on achievement, whereas it does
not play a role in later years of education (Laidra et al.,
2007).
Most studies have found a negative relation between
neuroticism and academic achievement (Poropat, 2009).
If a student is very anxious, this might interfere with his
attention to academic tasks and lead to poorer performance
(De Raad & Schouwenburg, 1996). A trait which specifi-
cally addresses this issue is test anxiety. It has been shown
that test anxiety is negatively correlated with academic
achievement at different educational levels (Hembree,
1988; Rindermann & Neubauer, 2001).
In addition to the discussed personality traits, we also
considered motivational variables like self-efficacy and
intrinsic versus extrinsic school motivation. Self-efficacy
is defined as ‘‘beliefs in one’s capabilities to mobilize the
motivation, cognitive resources, and courses of action
needed to meet given situational demands’’ (Wood &
Bandura, 1989, p. 408). This trait plays a role in various life
areas; among others, it is associated with job search success
(Saks, 2006), career success (Abele & Spurk, 2009), and
academic achievement (Caprara, Vecchione, Alessandri,
& Barbaranelli, 2011). Intrinsic-extrinsic motivation is
described in the framework of Deci and Ryan’s (1985)
self-determination theory. According to this theory, motiva-
tion is not only two-dimensional, but extrinsic motivation
can be divided again into the components external regula-
tion,introjected regulation,andidentified regulation.
Intrinsic regulation represents the highest level of
self-determination, external regulation the lowest. Self-
determined motivation turned out to be positively related
to educational outcomes (Deci, Vallerand, Pelletier, &
Ryan, 1991).
Most studies on the prediction of academic achievement
can be found in higher education (cf. Poropat, 2009). But if
we want to support students early in improving their
achievement to open them new possibilities for future
career, more research in earlier years of education is
needed. In the present study, we include all of the described
variables to find the best set of predictors for the school
achievement of 13–14-year-old adolescents. In contrast to
most of the studies mentioned above, which only focus
on certain personality or motivational traits, our aim was
to consider almost all variables that are often examined in
the context of school achievement. It could be possible that
some of the discussed variables turn out as redundant when
simultaneously assessed with other (similar) traits.
1
We did not only include various predictors of school
achievement simultaneously, but also examined their
impact on different criteria of school achievement, namely
GPA, and science versus languages. In this, we tested the
predictive power of the broad personality traits of the big
five but, additionally, also whether the – in this context
most relevant narrow personality traits – allow for an incre-
mental prediction of the various school achievement indica-
tors. The results of this study should allow us to conclude
which personality and motivational traits are most
important for students who want to improve their achieve-
ment and should, therefore, be assessed in counseling
contexts.
First, the relationships between school achievement and
the selected personality and motivational variables shall be
studied. Based on previous findings, we expect substantial
relationships of the broad big five factors conscientiousness
and openness with school achievement (Laidra et al., 2007;
Poropat, 2009). Furthermore, we assume relationships with
the narrow traits self-discipline, grit, test anxiety, self-
efficacy, and intrinsic (versus extrinsic) motivation (Caprara
et al., 2011; Deci et al., 1991; Duckworth et al., 2007;
Duckworth & Seligman, 2005; Hembree, 1988).
Second, the relative importance of the personality and
motivation predictors compared to intelligence should be
examined. Due to the characteristics of our sample that is
not restricted in intellectual range, intelligence should have
the highest impact on school achievement (Chamorro-
Premuzic & Furnham, 2005; Jensen, 1980). In this, we will
also explore the question if any of the selected personality
and motivational variables shows incremental validity
above the hitherto most potent noncognitive predictor of
school achievement, that is, conscientiousness. On the basis
of the findings of Paunonen and Ashton (2001) as well as
Lounsbury, Sundstrom, Loveland, and Gibson (2003) we
expect all narrow traits to potentially improve the
prediction.
Third, similarities and discrepancies in the prediction of
different criteria of school achievement – grade point aver-
age (GPA), science,andlanguages – shall be assessed.
Based on the findings of Furnham et al. (2011), intelligence
should be most important for science, whereas personality
and motivational traits should be most important for
languages.
1
Due to economic reasons, we could not include each relevant construct. In smaller pilot studies we found out that, for example, typical
intellectual engagement could not predict school achievement in a sample of average-achieving adolescents. Therefore, this trait was
omitted for the main study presented here.
B. Dumfart & A. C. Neubauer: Prediction of School Achievement in Adolescents 9
Author’s personal copy (e-offprint)
2016 Hogrefe Publishing Journal of Individual Differences 2016; Vol. 37(1):8–15

Method
Participants
We tested 498 students from general secondary schools.
This school track is more work-oriented (in contrast to aca-
demic secondary schools that prepare for university) and
attended by approximately two-thirds of Austrian adoles-
cents (Freudenthaler, Spinath, & Neubauer, 2008). General
secondary schools are open for all children independent of
previous school achievement. Due to incomplete data
regarding school grades, 137 persons had to be excluded
from the present analyses. The final sample consisted of
171 girls and 190 boys with a mean age of 14.09 years
(SD = 0.48). All students participated voluntarily with in-
formed consent of their parents. Cognitive abilities were
tested at the end of seventh grade under supervision of
trained testers whereas all other measures were taken a
few months later (under supervision of specially trained
teachers).
While school achievement, intelligence, the Big Five,
self-discipline, and self-efficacy were assessed in the entire
sample, the remaining personality and motivational vari-
ables were only tested within smaller subsamples. Grit
was measured in a sample of 129 persons (78 girls), test
anxiety in a sample of 131 persons (49 girls), and intrin-
sic-extrinsic motivation in a sample of 94 persons (40 girls).
Measures and Procedure
To obtain a measure for school achievement, students were
asked to report all grades of their last certificate (end of 7th
grade). For the GPA, we computed a weighted mean score
consisting of following grades: German, English, Math,
Physics, Biology, Geography, and History. German,
English, and Math were double-weighted as the curriculum
demands twice as many credit hours for these subjects. For
obtaining a measure of science, a weighted mean score con-
sisting of Biology, Physics, and Math (with a double weight
on math) was computed. For languages, the mean of
German and English was computed.
For intelligence, three subscales of the German test
‘‘Intelligenz-Struktur-Analyse’’ (ISA; Blum et al., 1998)
were used (verbal intelligence: ‘‘Gemeinsamkeiten finden’’
– commonalities; numeric intelligence: ‘‘Zahlenreihen
fortsetzen’’ – number series; visuospatial intelligence:
‘‘Figuren zusammensetzen’’ – composition of figures).
The time-restricted subscales comprise 20 items each and
show internal consistencies between .80 and .89. An EFA
of the subscales indicated a first unrotated factor with an
eigenvalue of 1.78 accounting for 41% of the variance.
For obtaining a measure of general intelligence, the factor
scores for the first unrotated factor were used for further
analyses.
To assess the Big Five, the German version of the
Hierarchical Personality Inventory for Children (HiPIC;
Mervielde & De Fruyt, 2002; German version by Bleidorn
& Ostendorf, 2009) was used. It comprises 144 items on
five factors, namely Neuroticism (N), Extraversion (E),
Imagination (I), Benevolence (B), and Conscientiousness
(C), on a 5-point Likert-type scale (responses range from
barely characteristic of me to highly characteristic of me).
The factors I and B can be regarded as equivalent to the
Big Five factors openness and agreeableness. The inventory
shows high reliabilities (a= .83–.88) in adolescents.
For measuring self-discipline and general self-efficacy,
new scales were developed that should be better suited
for the target sample of 13-year-olds, because the existing
questionnaires seem too difficult in verbal formulations to
be understood by 13-year-old students encompassing the
full range of verbal abilities. To generate the new items,
we extracted contents from well-established questionnaires
measuring related traits (e.g., NEO-PI-R; Costa & McCrae,
1992; HEXACO-PI-R; Lee & Ashton, 2004; Self-Control
Scale; Tangney, Baumeister, & Boone, 2004; general self-
efficacy [Allgemeine Selbstwirksamkeitserwartung];
Jerusalem & Satow, 1999) and reformulated the contents
in much simpler ways than in the classical questionnaires
that are targeted at adult persons. To ensure that the new
items are adequate for our adolescent sample, we conducted
cognitive interviews with the students as well as pilot
studies.
In their final versions, the self-discipline scale was com-
prised of six items, the self-efficacy scale of seven items.
We used a 4-point Likert-type scale ranging from not
appropriate to very appropriate. The results of the
confirmatory factor analyses indicated one-dimensionality
for each scale (self-discipline: v
2
(df =9)
=25.17,
p(Bollen-Stine Bootstrap) = .040; RMSEA = .048, CFI =
.97, SRMR = .029; self-efficacy: v
2
(df =14)
=31.22,
p(Bollen-Stine Bootstrap) = .040; RMSEA = .040,
CFI = .98, SRMR = .040). The internal consistencies (for
the present sample) are acceptable (self-discipline:
a= .63; self-efficacy: a= .72). The validity of the new
scales had been tested in pilot studies. For self-discipline,
construct validity was demonstrated by high correlations
with the facet scales concentration (r= .55) and persever-
ance (r= .60) of the Big Five factor conscientiousness
(assessed using the HiPIC; Mervielde & De Fruyt, 2002),
as well as by either low or moderate correlations with the
other HiPIC factors (neuroticism: r=.34, extraversion:
r= .12, imagination: r= .25, benevolence: r=.41) and
intrinsic motivation (r= .31; measured by the Academic
Self-Regulation Questionnaire SRQ-A; Ryan & Connell,
1989; German version by Müller, Hanfstingl, & Andreitz,
2007). Criterion validity could be shown by positive rela-
tionships with time spent on learning (r= .12), time of
day when homework is begun (r=.27) and negative rela-
tionships with watching TV (r=.14) and playing com-
puter games (r=.21; measured by asking for the
weekly time amount spent on certain activities). For self-
efficacy, we found relationships which were similar to those
of the self-efficacy scale by Jerusalem and Satow (1999),
for example, high negative relationships with the HiPIC
factor neuroticism (r=.50) and the scale lack of confi-
dence of the Test-Anxiety Questionnaire PAF (r=.54;
‘‘Prüfungsangstfragebogen’’; Hodapp, Rohrmann, &
Ringeisen, 2011).
10 B. Dumfart & A. C. Neubauer: Prediction of School Achievement in Adolescents
Author’s personal copy (e-offprint)
Journal of Individual Differences 2016; Vol. 37(1):8–15 2016 Hogrefe Publishing

To assess grit, the Short-Grit-Scale (Duckworth &
Quinn, 2009), which consists of eight items, was translated
into German. We used a 4-point Likert-type scale ranging
from not appropriate to very appropriate. The internal con-
sistency in our sample was sufficient (a= .70, N= 129).
Test anxiety was measured by the German test-anxiety
questionnaire PAF (Hodapp et al., 2011). The questionnaire
comprises of 20 items and shows high internal consistency
(a=.82–.90).
Intrinsic-extrinsic motivation was assessed by a German
adaptation of the SRQ-A (Ryan & Connell, 1989). It con-
sists of 17 items on four scales: intrinsic,identified,intro-
jected,andexternal regulation on a 5-point Likert-type
scale ranging from very appropriate to not appropriate.
The items are originally formulated to refer to only one
school subject, but were used here to assess general scholas-
tic motivation (e.g., ‘‘I mainly learn or study in school be-
cause it’s fun’’). The internal consistencies for the German
version are satisfying (a= .67–.90). The self-determination
index (SDI) is computed as suggested by the authors using
the following formula: 2 ·intrinsic +identified intro-
jected 2·external.
Results
Descriptive statistics as well as correlations among the vari-
ables can be seen in Table 1. GPA as a composite of the
other two criteria is – of course – highly correlated with
them; interesting is only that science and languages still
shared 58% of variance. Gender was slightly associated
with GPA and languages (girls performed better). Age
was slightly negatively correlated with all criteria. Intelli-
gence was highly related to school achievement, especially
with GPA and science, somewhat lower with languages.
The correlations between the Big Five factors and school
achievement were rather similar for the three criteria;
conscientiousness and imagination were moderately, extra-
version and benevolence slightly correlated with school
achievement. Neuroticism was slightly negatively associ-
ated with school achievement. Grit and self-efficacy were
slightly to moderately positively associated with the criteria.
Test anxiety and intrinsic-extrinsic motivation were only
slightly related to science, but not to GPA and languages.
To examine the relative importance of all personality
and motivational predictors, we performed hierarchical
regressions for all criteria. All regressions had the same
structure; to control for sex and age, these variables were
introduced in the first step. Due to the hypothesis that intel-
ligence and conscientiousness are the most important
predictor variables, they were included in the second step.
In the third step, additional personality or motivational pre-
dictors were included to assess incremental variance of
these variables. Since grit, test anxiety, and intrinsic-
extrinsic motivation were not tested within the entire
sample, but each of them within a subsample, not all pre-
dictors could be entered in one regression. To examine if
the subsamples for grit, extrinsic-intrinsic motivation, and
test anxiety are comparable to the larger sample, we
performed the main regression analyses (regressions 1a–c)
again within the subsamples. The results deviated only mar-
ginally with respect to the beta weights, only some of them
did no longer reach significance, which could be explained
by the decreasing statistical power of smaller samples.
The results, summarizing total R
2
after step 1 and
R
2
change after step 2, as well as the regression coefficients
of the final model, can be found in Tables 2–4. In none of
the regression analyses, the predictor variables showed
multicollinearity. The final regression models of the entire
Table 1. Descriptive statistics and correlations among the variables
NMSD 12345678910111213
1. Sex 361 – – –
2. Age 358 14.09 0.48 .01 –
3. GPA 327 3.59 0.90 .14* .16** –
4. Science 361 3.60 0.91 .09 .15** .94** –
5. Lang 339 3.39 1.01 .15** .18** .92** .76** –
6. IQ 361 0.00 0.85 .18** .14* .55** .56** .46** –
7. C 359 3.40 0.48 .02 .03 .34** .31** .28** .16** –
8. N 357 2.65 0.58 .28** .03 .16** .14** .13* .28** .37** –
9. E 358 3.36 0.47 .21** .07 .18** .11* .18** .10* .31** .26** –
10. I 359 3.41 0.51 .08 .02 .31** .27** .29** .28** .54** .33** .57** –
11. B 357 3.41 0.39 .25** .04 .19** .20** .14* .00 .40** .17** .16** .19** –
12. Self-d 360 7.08 1.39 .05 .00 .23** .25** .13* .04 .59** .34** .12* .25** .41** –
13. Self-e 358 20.19 3.20 .08 .02 .23** .22** .17** .24** .47** .50** .34** .53** .20** .32** –
14. Grit 129 19.04 3.37 .09 .05 .29** .20* .23* .15 .67** .54** .21* .39** .39** .67** .61**
15. TA 131 45.66 7.75 .21* .05 .14 .19* .08 .29** .24** .65** .05 .14 .05 .12 .28**
16. SDI 94 10.64 11.03 .18 .30** .19 .22* .12 .05 .40** .32** .14 .16 .28** .43** .29**
Notes. lang = languages; self-d = self-discipline; self-e = self-efficacy; TA = test anxiety; female = 1, male = 2. *p< .05. **p< .01
(two-tailed). Variables 14–16 could not be correlated because of non-overlapping samples.
B. Dumfart & A. C. Neubauer: Prediction of School Achievement in Adolescents 11
Author’s personal copy (e-offprint)
2016 Hogrefe Publishing Journal of Individual Differences 2016; Vol. 37(1):8–15

sample (regressions 1a–c; Table 2) accounted for 45% of
the variance in GPA and for 32% of the variance in lan-
guages. In the prediction of GPA and languages, only steps
1 and 2 (IQ, C) contributed significantly. The remaining
Big Five factors, as well as self-discipline and self-efficacy
(step 3), did not explain incremental variance. Intelligence
was the strongest predictor, explaining 26% of unique var-
iance in GPA and 16% in languages. Only in science, one of
the step 3 variables (self-discipline) could marginally en-
hance the prediction, uniquely explaining 1% of the vari-
ance. Intelligence was the strongest predictor, explaining
34% of unique variance.
In subsample 1 (regressions 2a–c; Table 3), the incre-
mental contribution of grit over and above intelligence
and conscientiousness was examined. Similar to the find-
ings of the full sample, the amount of prediction could only
be increased significantly in steps 1 and 2. Adding grit in
the third step could not enhance the prediction, either for
GPA, or for science or languages.
Due to nonsignificant zero-order correlations between
the predictors test anxiety and intrinsic-extrinsic motivation
with the criteria GPA and languages, regressions for these
traits were only performed for science. In subsample 2
(regression 3; Table 4), we analyzed the incremental vari-
ance of test anxiety by adding it in the third step. It could
not increase the amount of prediction. In subsample 3
(regression 4; Table 4), intrinsic-extrinsic motivation was
entered in the third step of the regression and could also
not account for incremental variance.
2
Discussion
We examined the impact of specific personality and moti-
vational predictors on the school achievement of eighth-
graders comparing three criteria: GPA (average across all
subjects), science subjects, and language subjects.
First, we deal with the bivariate correlations between
school achievement and all personality and motivational
predictors: The correlation between intelligence and school
achievement was high; compared to previous findings
(Furnham et al., 2011) it was considerably higher with sci-
ence than with languages. Although the three criterion vari-
ables were strongly intercorrelated, we found different
patterns of correlates with personality and motivational
traits. All Big Five traits as well as grit and self-efficacy
showed the highest associations with GPA. This could eas-
ily be explained by the presumed higher reliability of this
criterion compared to the others due to the higher level of
aggregation. By contrast, two of the narrow traits, namely
test anxiety and intrinsic-extrinsic motivation, correlated
significantly only with science. Comparable results could
be found in Spinath, Freudenthaler, and Neubauer (2010):
neuroticism showed incremental validity over intelligence
and ability self-perceptions only in Math achievement, not
in languages. Steinmayr and Spinath (2007) obtained simi-
lar results and conjectured that test anxiety seems to be
more important in domains where the correctness of an-
swers can be logically determined more easily, as it is the
case with Math. These findings could be interpreted in a
Table 2. Regressions 1a-c (r1a-r1c): Predicting school achievement by age, gender (step 1), intelligence,
conscientiousness (step 2), the remaining Big Five, self-discipline, and self-efficacy (step 3)
GPA (r1a, N= 318) Science (r1b, N= 351) Languages (r1c, N= 330)
R
2
DR
2
btR
2
DR
2
btR
2
DR
2
bt
Step 1 .05 .03 .06
F(2, 315) = 8.18** F(2, 348) = 6.07** F(2, 327) = 10.34**
Step 2 .44 .39 .41 .38 .32 .26
DF(2, 313) = 107.48** DF(2, 346) = 110.18** DF(2, 325) = 62.05**
Step 3 .45 .01 .43 .02 .32 .00
DF(6, 307) = 0.80 DF(6, 340) = 2.19* DF(6, 316) = 0.22
Sex .21 4.31** .15 3.26** .22 4.07**
Age .09 2.16* .08 1.85 .13 2.67**
IQ .56 12.06** .58 12.91** .44 8.70**
C .18 2.85** .15 2.50* .19 2.77**
N .06 1.09 .10 1.80 .01 0.24
E .00 0.08 .05 0.89 .01 0.08
I .01 0.20 .00 0.05 .06 0.89
B .04 0.88 .06 1.33 .01 0.25
Self-d .09 1.63 .13 2.30* .01 0.21
Self-e .02 0.29 .04 0.66 .02 0.44
Notes. self-d = self-discipline; self-e = self-efficacy. *p< .05. **p< .01 (two-tailed).
2
All regression analyses were repeated entering the narrow traits (self-discipline, self-efficacy, grit, test anxiety, and intrinsic-extrinsic
motivation) in the second step and conscientiousness in the third step. The results indicated significant contributions of all narrow traits in
the second step which were reduced to nonsignificance when adding conscientiousness in the following step. Conscientiousness
accounted for incremental variance over and above the narrow traits (for space constraints these results are not presented here but can be
obtained from the authors).
12 B. Dumfart & A. C. Neubauer: Prediction of School Achievement in Adolescents
Author’s personal copy (e-offprint)
Journal of Individual Differences 2016; Vol. 37(1):8–15 2016 Hogrefe Publishing

way that strategies to encourage the intrinsic motivation and
reduce the test anxiety of students are of particular impor-
tance in Math, maybe generally in science subjects.
Most bivariate correlations between school achievement
and the predictor variables correspond to previous findings
(Chamorro-Premuzic & Furnham, 2005; De Raad &
Schouwenburg, 1996; Poropat, 2009): moderate relations
with conscientiousness, imagination (openness), self-disci-
pline, grit, and self-efficacy as well as weak relations with
benevolence (agreeableness), test anxiety, and intrinsic-
extrinsic motivation. Contrary to previous findings, extra-
version was also slightly positively correlated with school
achievement, especially with GPA and languages. Spinath
et al. (2010) obtained similar results in eighth-graders
(but only in girls) which they explained by the higher
importance of oral performance in languages. If participat-
ing actively during class is requested, extraverted students
may have advantages in these subjects.
The hierarchical regressions indicated that all selected
predictors together explained almost half of the variance
in adolescent school achievement. Intelligence was by far
the most important predictor; although conscientiousness
could uniquely contribute to the prediction, the impact of
intelligence was much higher. This result is comparable
to findings of studies in similar populations; in nonselective
schools where intelligence is not (yet) range-restricted
(Freudenthaler et al., 2008; Laidra et al., 2007).
With respect to the GPA and languages, girls performed
significantly better than boys while there was no difference
in sciences. This is in line with previous findings (Freudent-
haler et al., 2008). Age had a negative impact on school
achievement. This may seem counterintuitive, but it can
be explained by the fact that all students were basically
from the same age cohort; most students who were older
than the average have likely repeated at least one grade
because of poor performance.
Although most bivariate correlations were as high
as expected on the basis of the literature, the third step
of the hierarchical regressions showed that – in addition
to conscientiousness – no other personality or motivational
traits could substantially enhance the prediction. Self-
discipline could account for only 1% of incremental
Table 3. Regressions 2a–c (r2a–r2c): Predicting school achievement in subsample 1 by age, gender (step 1), intelligence,
conscientiousness (step 2), and Grit (step 3)
GPA (r2a, N= 107) Science (r2b, N= 126) Languages (r2c, N= 114)
R
2
DR
2
btR
2
DR
2
btR
2
DR
2
bt
Step 1 .08 .08 .09
F(2, 104) = 4.25** F(2, 123) = 5.10** F(2, 111) = 5.75**
Step 2 .48 .41 .44 .36 .40 .30
DF(2, 102) = 40.23** DF(2, 121) = 39.16** DF(2, 109) = 27.39**
Step 3 .49 .00 .44 .00 .40 .00
DF(1, 101) = 0.55 DF(1, 120) = 0.02 DF(1, 108) = 0.43
Sex .21 2.80** .11 1.63 .25 3.34**
Age .15 1.97 .16 2.22* .17 2.16*
IQ .53 6.76** .53 7.04** .47 5.74**
C .23 2.20** .21 2.25* .17 1.57
Grit .08 0.74 .01 0.14 .07 0.65
Note. *p< .05. **p< .01 (two-tailed).
Table 4. Regressions 3 and 4 (r3, r4): Predicting Science by age, gender (step 1), intelligence, conscientiousness (step 2),
and test anxiety (step 3; subsample 2), respectively, intrinsic-extrinsic motivation (step 3; subsample 3)
Science (r3, N= 94) Science (r4, N= 130)
R
2
DR
2
btR
2
DR
2
bt
Step 1 .05 Step 1 .00
F(2, 127) = 3.23** F(2, 91) = 0.03
Step 2 .49 .44 Step 2 .29 .28
DF(2, 125) = 53.59** DF(2, 89) = 17.67**
Step 3 .49 .00 Step 3 .31 .02
DF(1, 124) = 0.03 DF(1, 88) = 3.03
Sex 0.25 3.78** Sex 0.09 0.94
Age 0.01 0.08 Age 0.05 0.55
IQ 0.63 9.04** IQ 0.48 5.19**
C 0.26 3.92** C 0.15 1.49
Test anxiety 0.01 0.18 SDI 0.18 1.74
Note. **p< .01 (two-tailed).
B. Dumfart & A. C. Neubauer: Prediction of School Achievement in Adolescents 13
Author’s personal copy (e-offprint)
2016 Hogrefe Publishing Journal of Individual Differences 2016; Vol. 37(1):8–15

variance over intelligence and conscientiousness in the pre-
diction of science; none of the narrow traits could enhance
the prediction of GPA or languages. Broad traits could sub-
stantially improve the prediction over and above narrow
traits, but not vice versa. That implies that in this age range
tested here there seems to be little benefit of assessing nar-
row traits like self-discipline or grit for counseling contexts.
For the target population of average-achieving adolescents
it would be sufficient to examine broad personality traits,
particularly conscientiousness, when supporting them in
improving school achievement. The narrower traits might
be of relevance in higher age ranges or more selective
schools like academic high schools. It may also be possible
that the broad traits outperformed the narrow traits because
of the broad performance criteria used in this study. It has
been shown that narrow traits are useful when predicting
specific performance criterion, as for instance, counterpro-
ductive work behaviors. However, when predicting a global
criteria, like overall job performance, broad traits are
preferable (Ones & Viswesvaran, 1996; Dudley, Orvis,
Lebiecki, & Cortina, 2006). Therefore, it would be interest-
ing to extend our study to additional performance criteria in
school, for example, class participation or oral exam
performance.
In summary, intelligence and conscientiousness turned
out – once again – as the most powerful predictors for
school achievement of adolescents. No other personality
and motivational variables (the remaining Big Five, self-
discipline, grit, self-efficacy, intrinsic-extrinsic motivation,
and test anxiety) could substantially enhance the prediction,
although most of them were correlated with grades.
As a restriction of our study it should be mentioned that
on the basis of a thorough literature research we included
only the most discussed variables with respect to the predic-
tion of school achievement. Due to economic reasons, we
could not test each variable which has turned out to be asso-
ciated with school achievement but focused on those which
turned out to be of importance in samples of comparable
age ranges and school types. For future directions, it would
be interesting to explore the impact of certain other con-
structs over and above personality and motivation. For
example, it would be interesting to also include approaches
to learning. Previous research demonstrated interactions
between personality traits and approaches to learning
regarding academic achievement (Diseth, 2003).
Based on our results, we conclude that conscientious-
ness is the crucial noncognitive trait in school achievement
of adolescents. Some authors suggest that high conscien-
tiousness can even compensate for low (fluid) intelligence;
that is, students with lower intelligence and high conscien-
tiousness could perform as well as their more intelligent
colleagues who do not show such structured and persever-
ing learning habits (Moutafi, Furnham, & Paltiel, 2004;
Wood & Englert, 2009). So it might be useful to focus on
the impact of conscientiousness in improving school
achievement, for example, for school career counselors.
Some conscientious behaviors – like being on time, tidying
up the workplace, or keeping focused on a task – can be
trained with little effort but might have considerable influ-
ence on school achievement.
References
Abele, A. E., & Spurk, D. (2009). The longitudinal impact of
self-efficacy and career goals on objective and subjective
career success. Journal of Vocational Behavior, 74, 53–62.
Bleidorn, W., & Ostendorf, F. (2009). Ein Big Five-Inventar für
Kinder und Jugendliche. Die deutsche Version des Hierar-
chical Personality Inventory for Children (HiPIC) [A Big
Five inventory for children and adolescents: The German
version of the Hierarchical Personality Inventory for Chil-
dren (HiPIC)]. Diagnostica, 55, 160–173.
Blum, F., Didi, H. J., Fay, E., Maichle, U., Trost, G., Wahlen,
J. H., & Gittler, G. (1998). Intelligenz Struktur Analyse
(ISA). Ein Test zur Messung der Intelligenz [Intelligence
Structure Analysis. A test for measuring intelligence].
Frankfurt, Germany: Swets & Zeitlinger B. V., Swets Test
Services.
Boekaerts, M. (1995). The interface between intelligence and
personality as determinants of classroom learning. In D. H.
Saklofske & M. Zeidner (Eds.), Handbook of Personality
and Intelligence (pp. 161–183). New York, NY: Plenum.
Caprara, G. V., Vecchione, M., Alessandri, M. G., &
Barbaranelli, C. (2011). The contribution of personality
traits and self-efficacy beliefs to academic achievement:
A longitudinal study. The British Journal of Educational
Psychology, 81, 78–96.
Chamorro-Premuzic, T., & Furnham, A. (2005). Personality and
intellectual competences. Mahwah, NJ: Erlbaum.
Costa, P. T. Jr., & McCrae, R. R. (1992). Revised NEO
personality inventory (NEO-PI-R) and NEO five-factor
inventory (NEO-FFI): Professional manual. Odessa, FL:
Psychological Assessment Resources.
Deary, I. J., Strand, S., Smith, P., & Fernandes, C. (2007).
Intelligence and educational achievement. Intelligence, 35,
13–21.
Deci, E. L., & Ryan, R. M. (1985). Intrinsic motivation and self-
determination in human behavior. New York, NY: Plenum.
Deci, E. L., Vallerand, R. J., Pelletier, L. G., & Ryan, R. M.
(1991). Motivation and education: The self-determination
perspective. The Educational Psychologist, 26, 325–346.
De Raad, B., & Schouwenburg, H. (1996). Personality in
learning and education: A review. European Journal of
Personality, 10, 406–425.
Diseth, Å. (2003). Personality and approaches to learning as
predictors of academic achievement. European Journal of
Personality, 17, 143–156.
Duckworth, A. L., Peterson, C., Matthews, M. D., & Kelly,
D. R. (2007). Grit: Perseverance and passion for long-term
goals. Journal of Personality & Social Psychology, 92,
1087–1101.
Duckworth, A. L., & Quinn, P. D. (2009). Development and
validation of the Short Grit Scale (GRIT–S). Journal of
Personality Assessment, 91, 166–174.
Duckworth, A. L., & Seligman, M. E. P. (2005). Self-discipline
outdoes IQ in predicting academic performance of adoles-
cents. Psychological Science, 16, 939–944.
Dudley, N. M., Orvis, K. A., Lebiecki, J. E., & Cortina, J. M.
(2006). A meta-analytic investigation of conscientiousness
in the prediction of job performance: Examining the
intercorrelations and the incremental validity of narrow
traits. Journal of Applied Psychology, 91, 40.
14 B. Dumfart & A. C. Neubauer: Prediction of School Achievement in Adolescents
Author’s personal copy (e-offprint)
Journal of Individual Differences 2016; Vol. 37(1):8–15 2016 Hogrefe Publishing

Freudenthaler, H. H., Spinath, B., & Neubauer, A. C. (2008).
Predicting school achievement in boys and girls. European
Journal of Personality, 22, 231–245.
Furnham, A., Rinaldelli-Tabaton, E., & Chamorro-Premuzic, T.
(2011). Personality and intelligence predict arts and science
school results in 16 year olds. Psychologia, 54, 39–51.
Goff, M., & Ackerman, P. L. (1992). Personality-intelligence
relations: Assessment of typical intellectual engagement.
Journal of Educational Psychology, 84, 537–552.
Gustafsson, J.-E., & Undheim, J. O. (1996). Individual
differences in cognitive functions. In D. C. Berliner &
R. C. Calfee (Eds.), Handbook of educational psychology
(pp. 186–242). New York, NY: Prentice Hall.
Hembree, R. (1988). Correlates, causes, effects, and treatment of
test anxiety. Review of Educational Research, 58, 47–77.
Hodapp, V., Rohrmann, S., & Ringeisen, T. (2011). PA F –
Prüfungsangstfragebogen [Test anxiety questionnaire].
Göttingen, Germany: Hogrefe.
Jensen, A. R. (1980). Bias in mental testing. New York, NY: The
Free Press.
Jerusalem, M., & Satow, L. (1999). Schulbezogene Selbst-
wirksamkeitserwartung [School-related self-efficacy]. In R.
Schwarzer & M. Jerusalem (Eds.), Skalen zur Erfassung von
Lehrer- und Schülermerkmalen (pp. 15–16). Berlin,
Germany: Freie Universität Berlin.
Laidra, K., Pullmann, H., & Allik, J. (2007). Personality and
intelligence as predictors of academic achievement: A cross-
sectional study from elementary to secondary school.
Personality and Individual Differences, 42, 441–451.
Lee, K., & Ashton, M. C. (2004). Psychometric properties of the
HEXACO personality inventory. Multivariate Behavioral
Research, 39, 329–358.
Lounsbury, J. W., Sundstrom, E., Loveland, J. L., & Gibson,
L. W. (2003). Broad versus narrow personality traits in
predicting academic performance of adolescents. Learning
and Individual Differences, 14, 65–75.
Mervielde, I., & De Fruyt, F. (2002). Assessing children’s
traits with the Hierarchical personality inventory for chil-
dren. In B. De Raad & M. Perugini (Eds.), Big Five
assessment (pp. 129–142). Göttingen, Germany: Hogrefe.
Moutafi, J., Furnham, A., & Paltiel, L. (2004). Why is
conscientiousness negatively correlated with intelligence?
Personality and Individual Differences, 37, 1013–1022.
Müller, F. H., Hanfstingl, B., & Andreitz, I. (2007). Skalen zur
motivationalen Regulation beim Lernen von Schülerinnen
und Schülern: Adaptierte und ergänzte Version des Aca-
demic Self-Regulation Questionnaire (SRQ-A) nach Ryan &
Connell [Scales to assess motivational regulation during
students’ learning]. Wissenschaftliche Beiträge aus dem
Institut für Unterrichts- und Schulentwicklung Nr. 1.
Klagenfurt: Alpen-Adria-Universität.
O’Connor, M. C., & Paunonen, S. V. (2007). Big Five person-
ality predictors of postsecondary academic performance.
Personality and Individual Differences, 43, 971–990.
Ones, D. S., & Viswesvaran, C. (1996). Bandwidth-fidelity
dilemma in personality measurement for personnel selec-
tion. Journal of Organizational Behavior, 17, 609–626.
Paunonen, S. V., & Ashton, M. C. (2001). Big Five predictors of
academic achievement. Journal of Research in Personality,
35, 78–90.
Poropat, A. E. (2009). A meta-analysis of the five-factor model
of personality and academic performance. Psychological
Bulletin, 135, 322–338.
Rindermann, H., & Neubauer, A. C. (2001). The influence of
personality on three aspects of cognitive performance:
Processing speed, intelligence and school performance.
Personality and Individual Differences, 30, 829–842.
Ryan, R. M., & Connell, J. P. (1989). Perceived locus of
causality and internalization: Examining reasons for acting
in two domains. Journal of Personality and Social Psychol-
ogy, 57, 749–761.
Saks, A. M. (2006). Multiple predictors and criteria of job
search success. Journal of Vocational Behavior, 68,
400–415.
Spinath, B., Freudenthaler, H. H., & Neubauer, A. (2010).
Domain-specific school achievement in boys and girls
predicted by intelligence, personality and motivation.
Personality and Individual Differences, 48, 481–486.
Stankov, L., Lee, J., Luo, W., & Hogan, D. J. (2012). Confi-
dence: A better predictor of academic achievement than self-
efficacy, self-concept and anxiety? Learning and Individual
Differences, 22, 747–758.
Steinmayr, R., & Spinath, B. (2007). Predicting School
Achievement from Motivation and Personality. Zeitschrift
für Pädagogische Psychologie, 21, 207–216.
Tangney, J. P., Baumeister, R. F., & Boone, A. L. (2004). High
self-control predicts good adjustment, less pathology, better
grades, and interpersonal success. Journal of Personality, 72,
271–323.
von Stumm, S., Hell, B., & Chamorro-Premuzic, T. (2011).
Curiosity predicts academic performance. The hungry mind:
Intellectual curiosity as third pillar of academic perfor-
mance. Perspectives on Psychological Science, 6, 574–588.
Wood, P., & Englert, P. (2009). Intelligence compensation
theory: A critical examination of the negative relationship
between conscientiousness and fluid and crystallised intel-
ligence. The Australian and New Zealand Journal of
Organisational Psychology, 2, 19–29.
Wood, R., & Bandura, A. (1989). Impact of conceptions of
ability on self-regulatory mechanisms and complex decision
making. Journal of Personality and Social Psychology, 56,
407–415.
Date of acceptance: March 23, 2015
Published online: February 29, 2016
Aljoscha C. Neubauer
Dept. of Psychology
University of Graz
Maiffredygasse 12b
8010 Graz
Tel. +43 316 380–5124
E-mail aljoscha.neubauer@uni-graz.at
B. Dumfart & A. C. Neubauer: Prediction of School Achievement in Adolescents 15
Author’s personal copy (e-offprint)
2016 Hogrefe Publishing Journal of Individual Differences 2016; Vol. 37(1):8–15