Studies in Higher Education

ISSN: (Print) (Online) Journal homepage: www.tandfonline.com/journals/cshe20

Cognitive and non-cognitive predictors of

academic success in higher education: a large-
scale longitudinal study

Gyöngyvér Molnár & Ádám Kocsis

To cite this article: Gyöngyvér Molnár & Ádám Kocsis (17 Oct 2023): Cognitive and non-
cognitive predictors of academic success in higher education: a large-scale longitudinal study,
Studies in Higher Education, DOI: 10.1080/03075079.2023.2271513

To link to this article: https://doi.org/10.1080/03075079.2023.2271513

© 2023 The Author(s). Published by Informa

UK Limited, trading as Taylor & Francis
Group

Published online: 17 Oct 2023.

Submit your article to this journal

Article views: 2451

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=cshe20
STUDIES IN HIGHER EDUCATION
https://doi.org/10.1080/03075079.2023.2271513

Cognitive and non-cognitive predictors of academic success in

higher education: a large-scale longitudinal study
a
Gyöngyvér Molnár and Ádám Kocsisb
a
Institute of Education, MTA–SZTE Digital Learning Technologies Research Group, University of Szeged, Szeged,
Hungary; bDoctoral School of Education, MTA–SZTE Digital Learning Technologies Research Group, University of
Szeged, Szeged, Hungary

ABSTRACT ARTICLE HISTORY

How important are learning strategies or personal attributes for learning Received 21 July 2023
outside of domain-specific knowledge or twenty-first-century transversal Accepted 11 October 2023
skills when predicting academic success in higher education? To
KEYWORDS
address this question, we conducted a longitudinal study among 1,681 academic success; higher
students at one of the leading universities in Hungary. Students took education; cognitive and
four tests and questionnaires when they entered the university. They non-cognitive predictors;
measured knowledge from their previous studies (reading and domain-specific knowledge;
mathematics), twenty-first-century skills like problem-solving and non- 21st-century skills; learning
cognitive variables in learning strategies, motivation and attitudes. In strategies
the next five years, we followed their academic progress with the aim
of detecting cognitive and non-cognitive predictors of academic
success toward obtaining a degree and developing training
programmes for the most important predictors to boost the probability
of academic success. The overarching model, which was controlled for
students’ socio-economic background, determined 18.4% of the
variance in later academic success. Domain-specific knowledge from
previous studies and learning motivation emerged as the strongest
predictor, learning strategies proved to be the most stable of the other
mediating factors as an independent predictor, and twenty-first-century
transversal skills proved to be unimportant in earning a university
degree. Results highlight the importance of transforming higher
education so that it is consistent with the expectations of the twenty-
first-century labour market, a future-fit education in Hungary.

Introduction
Across the globe, one of the main objectives of higher education is to increase the percentage of
young people with university degrees (Diaz Lema et al. 2023). This requires actions that make it poss-
ible for students not only to access university, but also to successfully complete it (Roberts 2011). As a
result – beyond the issue of rapid technological development and its effect on the behaviour, traits
and expectations of current generations – we are facing a continuously changing higher education
environment.
In Hungary, where the present study was conducted, there were about 100,000 students in higher
education 30 years ago, mostly in full-time programmes. Twenty years after the millennium, there

CONTACT Gyöngyvér Molnár [email protected] Institute of Education, MTA–SZTE Digital Learning Tech-
nologies Research Group, University of Szeged, Petofi S. sgt. 32-34., 6723 Szeged, Hungary
© 2023 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group
This is an Open Access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License (http://
creativecommons.org/licenses/by-nc-nd/4.0/), which permits non-commercial re-use, distribution, and reproduction in any medium, provided the
original work is properly cited, and is not altered, transformed, or built upon in any way. The terms on which this article has been published
allow the posting of the Accepted Manuscript in a repository by the author(s) or with their consent.
2 G. MOLNÁR AND Á. KOCSIS

are approximately three times more, with one third of them being part-time students. In 1990, there
were 170,000 children enrolled in the first grade, while this number is 90,000 today because of the
population decline that has been part of the demographic changes in Europe. In 1990, 36% of the
students applying to higher education were admitted, while this rate is almost two times higher
today at 70%. That is, out of the significantly fewer students enrolled in public education, signifi-
cantly more students are entering higher education, and, in the context of lifelong learning, more
and more individuals are returning to higher education at a later stage in life, resulting in signifi-
cantly different students in higher education in many respects. It should also be noted that there
are vastly more international students at Hungarian universities than there were in 1990 (OECD
2019), there are far more non-state schools, and many more Hungarians are studying abroad. As a
consequence of these developments, the composition of students has been continuously changing
in the last few decades, not only in Hungary, but also internationally, including their expectations,
learning methods, knowledge, opportunities and abilities. On average, 31% of students dropped
out of tertiary education in 2010 (OECD 2010), while (OECD 2022) 32% of them failed to complete
their course in postsecondary public institutions by the end of the theoretical duration of the pro-
gramme in 2022.
In this constantly evolving environment, higher education institutions must make considerable
efforts – based on the latest research results – to reduce dropout rates and, at the same time,
boost admission rates while continuously improving completion rates, which are among the
lowest in the OECD (OECD 2017).
Academic success can be described as a complex web of factors that involve personal, academic,
organisational, pedagogical and social dimensions (e.g. Alyahyan and Düştegör 2020; Diaz Lema
et al. 2023) and interact with and influence each other (York, Gibson, and Rankin 2015). These pre-
dictors are important, as they provide insights into an individual’s potential for academic perform-
ance and success in higher education settings. In order to optimise and reduce the complexity of
this phenomenon, most studies focus on one type of predictor and the acquisition of the first
twenty credits as extremely important milestones during university studies (Clercq, Galand, and
Frenay 2017; Díaz et al. 2020). There is a lack of longitudinal studies that involve the multidimension-
ality of these predictors. To achieve this, extensive longitudinal research has been initiated at one of
the leading universities in Hungary to identify these predictors, with appropriate trainings in modifi-
able cognitive and non-cognitive factors, including competencies, which are highlighted as key in
the marketplace of the twenty-first-century. The results of this study build the bases of students’ per-
sonalised suggestions, feedback and offered targeted online and/or face-to-face training pro-
grammes and courses which should increase the probability of successful completion and
graduation.

Predictors of academic success in higher education

Several classifications of predictors of academic success exist (Alyahyan and Düştegör 2020; Van
Rooij, Jansen, and van de Grift 2018), but they can basically be divided into two parts: cognitive
and non-cognitive predictors.
The most commonly studied cognitive predictors are (1) grade point average (GPA) and credits
earned (ECTS) (e.g. Li and Wong 2019; Naaman 2021; Van Herpen et al. 2020) provided they are
used as an indicator of academic success, not as an outcome variable; (2) prior academic achieve-
ment (PAA) measured by grades and standardised tests (Richardson, Abraham, and Bond 2012; Wes-
trick et al. 2021), such as school-leaving or college admission exam results, which mostly assess
aptitude in areas such as mathematics, reading and writing, explaining 17–25% of its variance
(Musso, Rodríguez Hernández, and Cascallar 2020; Pinxten et al. 2015; Westrick et al. 2021); and
(3) cognitive abilities, including measures of intelligence, reasoning, critical thinking and problem-
solving, twenty-first-century skills, which influence a student’s capacity to understand complex con-
cepts and to engage in analytical thinking and problem-solving effectively, but they do not seem to
 STUDIES IN HIGHER EDUCATION 3

be significant predictors (e.g. Molnár et al. 2021; Pastén 2021; York, Gibson, and Rankin 2015; Zlatkin-
Troitschanskaia, Shavelson, and Kuhn 2015).
The most commonly monitored non-cognitive predictors are (1) socio-economic factors (SES),
which explain 9–23% of the variance of academic success (Musso, Rodríguez Hernández, and Cascal-
lar 2020; Rodríguez-Hernández et al. 2021); (2) learning and self-regulation strategies (Alhadabi and
Karpinski 2020; Farruggia et al. 2018; Musso, Rodríguez Hernández, and Cascallar 2020; Ribeiro et al.
2019), which are also mediating factors for GPA and ECTS and explain 16–26% of the variance of aca-
demic success (Aydin 2017; Bäulke, Eckerlein, and Dresel 2018); and (3) motivational factors, such as
goal orientation and self-efficacy (Alban and Mauricio 2019; Behr et al. 2020; Bowles and Brindle
2017; Ndoye, Clarke, and Henderson 2020; Rump, Esdar, and Wild 2017; Vanthournout et al. 2012),
which explain 9–20% of the variance of academic success (Alhadabi and Karpinski 2020; Azila-
Gbettor et al. 2021; Ndoye, Clarke, and Henderson 2020; Rump, Esdar, and Wild 2017), though
several studies indicate them as negative (Bäulke, Eckerlein, and Dresel 2018) or non-significant
factors (Alban and Mauricio 2019; Li and Wong 2019).

Information on the education system in Hungary: composite entry score and the
structure of the various study programmes
After high school, students take school-leaving, or Matura, examinations, whose results are part of
the entry score for higher education. From 2005, they consist of five subjects: a written examination
in mathematics, oral and written examinations in history, a foreign language, and Hungarian litera-
ture and grammar, and a written and/or oral examination in a subject of the student’s choice and can
be taken at an intermediate or advanced level. Study programmes in each higher education institute
determine the necessity of an advanced-level examination in particular subjects. School-leaving
exams are evaluated both by grade (from 1 to 5, with 5 being the best) and percentage.
The maximum value for the entry score is 500 and is calculated based on (1) academic results,
based on the average of the last two semester grades in Hungarian language and literature,
History, Mathematics, a foreign language and a natural science subject as well as school-leaving
examination results (max. 200 points), (2) school-leaving examination results in percentage form
in two subjects determined by the study programme (max. 200 points), and (3) extra points (max.
100 points, for example, for language skills, competition results, advanced-level examinations and
SES) or by doubling the school-leaving examination results with the extra points, which makes it
possible to compensate weak general knowledge with strong programme-specific knowledge. For
a detailed description of the system, see Molontay and Nagy (2023).
Hungary belongs to the European Higher Education Area and has a multi-cycle higher education
system with bachelor’s, master’s and postgraduate cycles in accordance with the Bologna process
since 1999. However, there are some exceptions, including dental and veterinary studies, law, medi-
cine, pharmacy and teacher training, which have a long, single-cycle study structure of five, five-and-
a-half or six years of study. Courses can be studied full-time, part-time or through distance learning.

Aims
The aim of the study is to map significant factors that influence students’ academic success in higher
education today. First, we monitor the predictive power of (a) domain-specific knowledge explicitly
taught during previous studies in secondary education (e.g. mathematics and reading); (b) cognitive
transversal skills, whose development is not explicitly pursued in secondary education but which
play a key role in the twenty-first-century (e.g. problem-solving and inductive reasoning); and (c)
non-cognitive factors, which can be modified with targeted training and is strongly connected to
the processes of learning: learning strategies, motivation and attitudes. In the second half of the
paper, we build an overarching model involving cognitive and non-cognitive factors to predict aca-
demic success. The research questions are:
4 G. MOLNÁR AND Á. KOCSIS

RQ1: To what extent do cognitive and non-cognitive factors each predict academic success?

RQ1a: To what extent does domain-specific knowledge from previous studies predict academic success in higher
education?

RQ1b: To what extent does the developmental level of twenty-first-century transversal skills predict academic
success in higher education?

RQ1c: To what extent do learning strategies, motivation and attitudes predict academic success at the university
level?

RQ2: To what extent does the overarching model of cognitive and non-cognitive factors predict academic
success at university? Which factors and what kind of knowledge, skills and attitudes are the most important?

Methods
Participants
Participants in the study were students admitted to a large Hungarian university, starting their
studies in the same year and followed in the next five years. Eleven faculties out of the twelve at
the university participated in the project. The faculties vary in number of students; that is, the per-
centages of students from the different faculties are not equally distributed in the sample, as they
mirror the distribution of the original population. The target population for the study was 3429 stu-
dents, of whom 1681 (49.02%) were involved in the analyses and 51% were female. Participants’
mean age was 19.86 (SD = 2.11), with most of them having taken their school-leaving examinations
in May of the same year. The first point of the data collection was in September, when they entered
the university.
Participation was voluntary, and the project was integrated into the educational improvements at
the university. Students were notified of the possibility of taking part in the assessment prior to start-
ing their studies. All the participants had turned 18 by the time of the assessment and confirmed
with their signature that their data could be used for educational and research purposes at both
the faculty and university levels. As an incentive, they received course credit for active participation
in the project. Annex 1 contains the descriptive statistics for the demographic variables of participat-
ing and non-participating students.

Instruments
Mathematics
Learning mathematics stimulates cognitive development, while mathematics provides knowledge and
skills that are essential for everyday tasks. It is one of the most important domains of education in every
school system. The test covered the major topics in mathematics learnt in secondary education. Its
difficulty level was adjusted approximately to the intermediate-level standards of the school-leaving
examination in mathematics. It measured two dimensions: disciplinary knowledge and its application.
The number of items and the reliability of the instrument are summarised in Table 1.

Reading comprehension
Reading comprehension is a gateway to learning and developing academic skills at every level of the
education system, especially in higher education, where students are required to read constantly and
reading encompasses different types of text (Pirttimaa, Takala, and Ladonlahti 2015). The reading
comprehension test involved four different types of text (narrative, explanatory, chart and Venn
diagram) and corresponding tasks in both continuous and discontinuous formats. Fifty-eight per
cent of the items were related to continuous texts, while two fifths measured understanding of dis-
continuous texts. The test was developed by Hódi and Tóth (2019). The number of items and the
reliability of the instrument are summarised in Table 1.
 STUDIES IN HIGHER EDUCATION 5

Table 1. Reliability of the tests and questionnaires and other descriptive statistics.
Test Number of items Cronbach’s alpha Min. Max. Mean SD
Mathematics 70 0.926 0 100 56.13 18.47
Reading comprehension 72 0.749 0 100 84.44 7.33
Problem-solving 24 0.892 0 100 53.97 22.67
Knowledge acquisition 12 0.865 0 100 60.53 27.20
Knowledge application 12 0.783 0 100 47.41 22.78
Inductive reasoning 48 0.868 0 100 65.83 15.45
Learning strategies, motivation and attitudes 78 0.916 1 5 3.56 0.35
Learning strategies Elaboration 4 0.663 1 5 3.63 0.63
Planning 8 0.955 1 5 3.70 0.68
Memorisation (rehearsal) 7 0.763 1 5 3.73 0.64
Procrastination 4 0.937 1 5 2.72 1.02
Help – support request 3 0.847 1 5 3.79 0.82
Time management 4 0.951 1 5 3.48 0.81
Learning motivation Mastery motivation 5 0.751 1 5 4.11 0.57
Achievement motivation 6 0.745 1 5 2.92 0.75
Learning attitudes Effort regulation 15 0.950 1 5 3.71 0.56
Cooperation 17 0.917 1 5 3.72 0.67
Openness to problem-solving 5 0.811 1 5 3.81 0.68

Problem-solving
Problem-solving plays an important role in twenty-first-century learning. It was measured via a
widely used validated instrument (see Greiff et al. 2013; Molnár, Ahmad Alrababah, and Greiff
2022; Molnár et al. 2017) containing computer-simulated problems developed within the MicroDYN
approach (Funke 2001). This type of problem-solving measure enables us to study how knowledge is
gathered in new situations and how this knowledge is applied to actually solving a problem, inde-
pendently of domain-specific content (Molnár, Greiff, and Csapó 2013). The number of items and the
reliability of the instrument are summarised in Table 1.

Inductive reasoning
Inductive reasoning is one of the most important reasoning skills with a central role in learning
(Hamers, Koning, and Sijtsma 1998; Molnár 2011) and is related to almost all higher-order cognitive
skills and processes (e.g. general intelligence, Klauer and Phye 2008; problem-solving, Molnár, Greiff,
and Csapó 2013; analogical reasoning, Goswami 1991). The inductive reasoning test was originally
developed by Csapó (1997) and further developed, computerised and validated at both the national
and international levels in the last 25 years (Mousa and Molnár 2020; Pásztor et al. 2018; Wu, Saleh,
and Molnár 2022; Wu and Molnár 2018). The current version contains figural analogies and series and
verbal analogies and series. Students were expected to discover the correct relationship between
given figures and numbers and select – using the drag-and-drop operation – a suitable figure or
number from among the five possibilities provided as their answer. The number of items and the
reliability of the instrument are summarised in Table 1.

Self-regulated learning and study-related resources: Learning strategies, motivation,

attitudes
Self-regulated learning strategies and study-related resources are used to assist students in learning
effectively; they entail the cognitive, metacognitive, behavioural, motivational and emotional/
affective aspects of learning (Panadero 2017). The learning strategies, motivation and attitudes ques-
tionnaire contains statements developed and adapted by D. Molnár and Gál (2019). The statements
were evaluated on a five-point Likert scale, ranging from 1 (strongly disagree) to 5 (strongly agree).
Based on Pintrich’s (1999) model of self-regulated learning, the statements for monitoring learning
strategies covered cognitive strategies, metacognitive awareness and resource management. State-
ments on cognitive strategies are tied to the selection, encoding and organisation of information
(e.g. memorisation and planning; Artelt et al. 2003; Weinstein, Husman, and Dierking 2000).
6 G. MOLNÁR AND Á. KOCSIS

Statements on metacognitive awareness and regulatory strategies pertain to the planning, monitor-
ing and evaluation of learning processes (e.g. elaboration and need for help; Adams and Blair 2019;
Winne and Perry 2000), while statements on resource management strategies encompass the man-
agement of external factors and internal resources (e.g. time management) (Pintrich 2000). State-
ments on learning motivation monitored factors that examine why students learn, why they start
and maintain their learning, and what they believe about their ability to learn subject matter. State-
ments on learning motivation involved mastery and achievement motivation (D. Molnár 2013).
Learning attitudes encompassed factors that measured students’ attitudes towards learning, such
as cooperation with peers (Kuger et al. 2016) and openness to different problem situations (D.
Molnár and Gál 2019). The analysis focused on areas where effective intervention and improvement
are possible. Table 1 summarises the areas covered in the analysis, the number of items per area,
reliability indices and other descriptive statistics.

Procedures
The assessments involving two testing sessions of two hours each were carried out in a large com-
puter room at the university’s Learning and Information Centre during the first two weeks of the
semester. The tests and questionnaires, measuring mathematics, reading comprehension,
problem-solving, learning strategies, motivation and attitudes and related background factors
(mothers’ education and number of books at home) were administered online with the eDia
online platform (Csapó and Molnár 2019). Students had a total of two times 120 min to complete
it. At the beginning of the testing sessions, participants were provided instructions on the user inter-
face, including warm-up tasks. Immediate average achievement-based feedback was provided after
they completed each of the tests along with detailed feedback with normative comparative data on
their performance via e-mail a week after the data collection was closed.
The general and follow-up educational data (e.g. entry score, length of studies, student’s faculty,
academic track, attempted course credits each semester, corrected credit index in each semester,
level of training – dropped out, still studying or degree received) were provided by the university’s
Office of the Director of Academic Affairs.
Data from the achievement tests were transformed to a 500(100) scale so that the university
means were set to 500 with Rasch scaling. MPlus software (Muthén and Muthén 2012) was employed
to conduct the structural equation modelling. A full SEM (structural equation model) was used to
investigate the roles of the cognitive, non-cognitive and demographic predictors on students’
later academic success. Academic success as a latent variable was defined by the corrected credit
index1 for the first and second semesters, by the global corrected credit index (GCCI) based on
the five-year follow-up and by the official status of the student (dropped out, still studying or
degree received) controlled for the length of the training. χ2 values, an absolute fit index (the
root mean square error of approximation, RMSEA) and two incremental fit indices (the Tucker–
Lewis Index, TLI, and the comparative fit index, CFI) were computed to evaluate model fit. A TLI
and CFI larger than 0.90 paired with a RMSEA less than 0.08 are commonly considered as an accep-
table model fit (Yu 2002). Modification indexes (MI) were used to add arrows to the model and
improve the model fit. Only significant coefficients were published in the SEM models.

Results
Mathematics, reading comprehension and previous studies are important factors in
predicting academic success
A full SEM model was used to explore the predictive power of previous studies for academic success
in higher education (see Figure 1). Academic success as a latent variable was defined by four mani-
fest variables: the corrected credit index for the first two semesters, GCCI and the official status of the
 STUDIES IN HIGHER EDUCATION 7

Figure 1. Domain-specific knowledge as a predictor of academic success in higher education controlled for students’ socio-econ-
omic background: The full SEM model.

student (see Procedures part). Knowledge from previous studies as a latent variable was defined by
the level of students’ reading comprehension, mathematics knowledge and entry score as manifest
variables. Because of the selectiveness of the Hungarian school system, which is mostly based on
parents’ education and socio-economic status (SES) (Csapó, Molnár, and Kinyó 2008; Csapó et al.
2014), we controlled domain-specific school knowledge as a latent factor of SES, which was built
on mother’s education and number of books at home. The controlled model fitted the data well
(χ2 = 338.51, CFI = .956, TLI = .929, RMSEA = .093, C.I. = .084.101) and explained 13.1% of the variance
of academic success. Previous school knowledge predicted later academic success at a medium level
(β = .362), while university entry score (mostly containing information about students’ GPA and
results on the school-leaving examinations) predicted student’s corrected credit index for the first
semester at a low level (β = .174). The first-semester corrected credit index correlated negatively
with the present status of the students after five years (r = −.259).

twenty-first-century transversal skills are basically not necessary for academic success in
the twenty-first century
Two highly correlated (r = .722 on a latent level) but distinct transversal skills that play an important
role in learning were assessed as a measure of twenty-first-century transversal skills, problem-solving
and inductive reasoning. A full SEM model was used to explore their predictive power on academic
success at a latent level (see Figure 2). The model had good model fits (χ2 = 144.609, CFI = .984, TLI

Figure 2. Twenty-first-century transversal skills as a predictor of academic success in higher education: The full SEM model.
8 G. MOLNÁR AND Á. KOCSIS

= .971, RMSEA = .069, C.I. = .059–.080). Problem-solving and inductive reasoning only explained 2.9%
of the variance of academic success, and the path coefficient was negative (β = −.247) for problem-
solving and positive but very low (β = .179) for inductive reasoning. A large amount of variance
remained unexplained.

Learning strategies, motivation and attitudes are predictive factors of academic success
at university level
The full SEM model (see Figure 3) fitted the data well (χ2 = 650.33, CFI = .953, TLI = .936, RMSEA
= .067, C.I. = .062–.072) and explained 16.8% of the variance of academic success. Not all the manifest
and latent factors proved to be significant. As a latent factor, learning strategies built on planning,
memorisation, elaboration, time management, asking for help and procrastination proved to have
the lowest but significant predictive power for later academic success (β = .220). As a latent factor,
learning motivation built on achievement motivation and mastery motivation proved to be the
most important predictor (β = .512) of later academic success. Finally, learning attitude at a latent
level defined by openness to problem-solving, cooperation and effort had a significant but negative
path as regards learning success (β = −.497).

The overarching model with cognitive and non-cognitive predictors of academic success
controlled for SES explains one fifth of academic success
The full SEM model (see Figure 4) fitted the data well (χ2 = 1404.85, CFI = .932, TLI = .913, RMSEA
= .058, C.I. = .055–.060) and explained 18.4% of the variance of academic success. (Please note
that, because of the high complexity of the full model, Figure 4 contains just the paths and
covariance arrows between the latent variables.) The latent variable of previously acquired,
domain-specific school knowledge defined by standardised test results in mathematics and
reading comprehension and students’ entry score proved to be the strongest predictor (β = .293)

Figure 3. Learning strategies, motivation and attitudes as predictors of academic success in higher education: The full SEM
model.
 STUDIES IN HIGHER EDUCATION 9

Figure 4. The overarching model contains domain-specific knowledge, twenty-first-century skills and learning strategies, motiv-
ation and attitudes as predictors of academic success in higher education: The full SEM model.

of academic success, while twenty-first-century transversal skills were not conducive to earning a
degree in higher education; however, they correlated strongly with previous school learning (r
= .500 for problem-solving and r = .690 for inductive reasoning). Specifically, students’
problem-solving skills negatively predicted their later academic success (β = −.167), and their
level of inductive reasoning proved to be a non-significant factor; however, it correlated strongly
with previous academic success (r = .690). Among non-cognitive factors, learning strategies (β
= .217) and learning motivation ( = .337) had positive predictive power for academic success
on a latent level, while learning attitude had negative predictive power (β = −.401). However,
the latter two constructs were strongly correlated (r = .806). Students’ entry score predicted
their quality achievement in the first semester at almost the same level as their level of inductive
reasoning skills (β = .171 and .111, separately). SES was not a significant influential factor at uni-
versity level.

Discussion
Prediction of academic success in higher education represents one of the major goals in psychologi-
cal and educational research in higher education. Because of the constantly changing circumstances
(e.g. demands of the marketplace, trends in technological developments and type of study) and par-
ticipants (e.g. in expectations, age and available opportunities), it requires continuous research.
Recent developments cannot be based on research results from a decade ago.
10 G. MOLNÁR AND Á. KOCSIS

The first purpose of this study was to investigate the relative contribution to academic success of
(1) domain-specific knowledge, that is, knowledge learnt and explicitly developed in secondary edu-
cation, (2) twenty-first-century transversal skills and (3) learning strategies, motivation and attitudes.
The factors proved to be predictive of academic success at different levels if measured alone,
accounting for 13.1, 2.9 and 16.8% of its variance, respectively.
We confirmed previous research results, which considered general and domain-specific knowl-
edge as a promising predictor, since students with a solid foundation in domain-specific knowledge
are better equipped to engage in the subject matter, understand it and earn higher grades. They
generally achieve higher results in the first semester (β = .172) and are more successful, indepen-
dently of the study programme (β = .362). Indeed, achieving good grades and accumulating an ade-
quate number of credits in the first semester can provide a sense of accomplishment and
reassurance, and they demonstrate that the student is capable of meeting the academic require-
ments of the programme. However, they do not guarantee continuous academic success and gradu-
ation at the end of the training (r = −.259). A large amount of variance remained unexplained
(86.9%), thus indicating that domain-specific and general disciplinary knowledge plays an important
role, but it is essential to recognise that academic success is multidimensional and influenced by
various other cognitive and non-cognitive factors.
Nowadays, many study programmes incorporate interdisciplinary approaches and provide edu-
cation that goes beyond narrow specialisation to ensure students are well-rounded and prepared
in a rapidly changing world. Thus, we expected that twenty-first-century transversal skills, especially
inductive reasoning and problem-solving, would have a high predictive power for academic success
in higher education. We detected a positive but very low predictive power for inductive reasoning in
academic achievement and a stronger but negative path for problem-solving. This means that stu-
dents equipped with higher-level twenty-first-century skills have generally no advantage at univer-
sity level; moreover, good problem-solvers have higher chances of dropping out than graduating in
the current system. The two constructs explained just 2.9% of the variance of academic success,
which also indicates that study programmes in Hungary should be further revised to better
respond to labour market demands and expectations and foster twenty-first-century skills.
Students reporting higher self-efficacy in learning are more successful, make more effort to com-
plete tasks (Olivier et al. 2019), have more control over their learning, and are generally more self-
confident (Luszczynska, Gutiérrez-Doña, and Schwarzer 2005; Martos et al. 2021). This means self-
regulated learning strategies and learning resources play an important role in learning success,
which was partially confirmed by our results. The learning strategies, motivation and attitudes we
monitored explained 16.8% of the variance of academic success, which is at about the same level
as the explained variance of domain-specific and general knowledge and which is significantly
larger than the predictive power of twenty-first-century skills. Learning motivation defined by
mastery and achievement motivation had a strong predictive power for final academic success,
and learning strategies proved to be important factors. Learning attitudes were determined by
effort, cooperation and openness to problem-solving; that is, constructs closely related to twenty-
first-century skills had a negative predictive power for later academic achievement.
The overarching model explained less than one fifth of the variance of academic success (18.4%)
and not one third of it, as expected by simply adding the results of the separately built models, indi-
cating the very complex phenomenon of academic success in higher education and the necessity of
monitoring it in a complex way, not just focusing on single domains of learning.
Learning motivation and domain-specific knowledge dominated as predictors; however, their pre-
dictive power dropped, including other cognitive (e.g. problem-solving) factors in the model. The pre-
dictive power of inductive reasoning became non-significant in the complex model; that is, other
learning-related cognitive and affective factors absorbed its predictive power. The most stable predic-
tive factor proved to be learning strategies, whose path coefficients remained almost the same in the
stand-alone and complex models, indicating their stable importance and necessity for development
consistent with academic success. Against our expectations, twenty-first-century skills, specifically
 STUDIES IN HIGHER EDUCATION 11

skills which make it possible to learn in unknown situations, emerged as almost unimportant factors in
academic success as well as learning attitudes, thus confirming earlier research results and highlight-
ing the importance of study programme revisions to better respond to the demands of the twenty-
first-century labour market. However, though not directly monitored, it is a very positive result of
the present analysis that SES proved not to be a significant influential factor at university level; that
is, differences brought from home are equalised at the level of higher education.

Limitations
The study sample may lead to limitations and generalisability of the results at a population level. We
used a non-representative convenience sample from one of the highest-ranked universities in
Hungary. We followed participants’ academic track record for five years. There are many participants
in the system because of the length of their training (e.g. medicine) or due to a changed study track.
We analysed all the students collectively, independently of their field of study and the expectations
of the different courses of study and some majors presumably rely on more problem-solving, critical
thinking and other twenty-first-century skills than others. We focused on problem-solving and induc-
tive reasoning within the scope of twenty-first-century skills, areas that only represent a subset of
that broad category. We focused on constructs which can be modified with specialised face-to-
face or online training.

Conclusion and further studies

This five-year longitudinal study investigated students’ cognitive and non-cognitive skills to monitor
their predictive power on later academic track and university graduation. Specifically, we monitored
students’ level of domain-specific knowledge on entering higher education, the level of twenty-first-
century transversal skills and the type of learning strategies, motivation and attitudes to ascertain the
causes behind a lack of successful completion of studies. The factors proved to be predictive of aca-
demic success at different levels if measured alone, thus accounting for 13.1, 2.9 and 16.8% of its
variance, respectively. A large amount of the variance remained unexplained in the overarching
model (81.6%), indicating the complexity of this phenomenon, the low level of predictability and
the high level of mediating roles of different factors, which can be examined with significant limit-
ations when separated from each other. Learning motivation and domain-specific knowledge
proved to be the most dominant predictors and learning strategies the most stable predictor. Con-
trary to our expectations, twenty-first-century transversal skills emerged as unimportant factors, thus
highlighting the significance of transforming higher education in accordance with the expectations
of the twenty-first-century labour market in Hungary, and success in the first semester does not guar-
antee continued academic success or graduation. Next, we will develop face-to-face and online
courses to facilitate students’ academic success in higher education in general and after division-
level analyses in particular in cooperation with the university’s education council.

Note
1. “Government Decree 76/2006. § 24(3) A student’s quantitative and qualitative academic performance during a
semester is measured by the credit index and the corrected credit index. The credit index is calculated by mul-
tiplying the credit point value of the subjects completed during the semester and the grades awarded to them,
which is then divided by the thirty credit points that indicate a student’s standard study progress during a seme-
ster. The corrected credit index is calculated by dividing the number of credit points attempted in a given seme-
ster by the number of credit points earned in the same semester, which is, in turn, multiplied by the value of the
credit index. The global corrected credit index is used to measure the student’s qualitative and quantitative aca-
demic performance over a span of multiple semesters. The global corrected credit index is calculated in the same
way as the corrected credit index except that it is thirty credit points per semester that have to be considered;
12 G. MOLNÁR AND Á. KOCSIS

moreover, the credit points attempted and subsequently earned are to be considered within the entire period.”
Source: Nftv: Act CCIV of 2011 on National Higher Education.

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

Funding
This work was supported by Hungarian Academy of Sciences Research Programme for Public Education Development
grant [grant number: KOZOKT2021-16]; Hungarian National Research, Development and Innovation Fund grant [grant
number: OTKA K135727 funding scheme].

ORCID
Gyöngyvér Molnár http://orcid.org/0000-0003-4890-6904

References
Adams, Richelle V., and Erik Blair. 2019. “Impact of Time Management Behaviors on Undergraduate Engineering
Students’ Performance.” Sage Open 9 (1): 2158244018824506.
Alban, Mayra, and David Mauricio. 2019. “Predicting University Dropout Trough Data Mining: A Systematic Literature.”
Indian Journal of Science and Technology 12 (4): 1–12. https://doi.org/10.17485/ijst/2019/v12i4/139729.
Alhadabi, Amal, and Aryn C. Karpinski. 2020. “Grit, Self-Efficacy, Achievement Orientation Goals, and Academic
Performance in University Students.” International Journal of Adolescence and Youth 25 (1): 519–35. https://doi.
org/10.1080/02673843.2019.1679202.
Alyahyan, Eyman, and Dilek Düştegör. 2020. “An Academic Arabic Corpus for Plagiarism Detection: Design, Construction
and Experimentation.” International Journal of Educational Technology in Higher Education 17: 1–21. https://doi.org/
10.1186/s41239-019-0174-x.
Artelt, Cordula, Jürgen Baumert, Nele Julius-McElvany, and Jules Peschar. 2003. “Learners for Life.” In Student Approaches
to Learning. Results from PISA 2000. Paris: OECD.
Aydin, Gokcen. 2017. “Personal Factors Predicting College Student Success.” Eurasian Journal of Educational Research 17
(69): 93–112. https://doi.org/10.14689/ejer.2017.69.6.
Azila-Gbettor, Edem M., Christopher Mensah, Martin K. Abiemo, and Marian Bokor. 2021. “Predicting Student
Engagement from Self-Efficacy and Autonomous Motivation: A Cross-Sectional Study.” Cogent Education 8 (1):
1942638. https://doi.org/10.1080/2331186X.2021.1942638.
Bäulke, Lisa, Nicole Eckerlein, and Markus Dresel. 2018. “Interrelations Between Motivational Regulation, Procrastination
and College Dropout Intentions.” Unterrichtswissenschaft 46 (4): 461–79. https://doi.org/10.1007/s42010-018-0029-5.
Bayer, Jaroslav, Hana Bydzovska, Jan Geryk, Tomas Obsivac, and Lubomir Popelinsky. 2012. “Predicting Drop-Out from
Social Behaviour of Students.” In EDM 2012 Proceedings of the Fifth International Conference on Educational Data
Mining, edited by Y. Kalina, Z. Osmar, H. Arnon, Y. Michael, and S. John, 103–9.
Behr, Andreas, Marco Giese, M. Herve, D. Teguim Kamdjou, and Katja Theune. 2010. Dropping Out of University: A
Literature Review.” Review of Education 8 (2): 614–52. https://doi.org/10.1002/rev3.3202.
Bowles, Terence V., and Kimberley A. Brindle. 2017. “Identifying Facilitating Factors and Barriers to Improving Student
Retention Rates in Tertiary Teaching Courses: A Systematic Review.” Higher Education Research & Development 36 (5):
903–19. https://doi.org/10.1080/07294360.2016.1264927.
Csapó, Benő. 1997. “The Development of Inductive Reasoning: Cross-Sectional Assessments in an Educational Context.”
International Journal of Behavioral Development 20 (4): 609–26.
Csapó, Benő, József B. Fejes, László Kinyó, and Edit Tóth. 2014. “Az iskolai teljesítmények alakulása Magyarországon
nemzetközi összehasonlításban.” 110–136.
Csapó, Benő, and Gyöngyvér Molnár. 2019. “Online Diagnostic Assessment in Support of Personalized Teaching and
Learning: The eDia System.” Frontiers in Psychology 10: 1522. https://doi.org/10.3389/fpsyg.2019.01522.
Csapó, Benő, Gyöngyvér Molnár, and László Kinyó. 2008. “Analysis of the Selectiveness of the Hungarian Educational
System in International Context.” IRC2008_Csapo_Molnar_etal. http://publicatio.bibl.u-szeged.hu/25905/1/
IRC2008_Program.pdf.
De Clercq, Mikaël, Benoît Galand, and Mariane Frenay. 2017. “Transition from High School to University: A Person-
Centered Approach to Academic Achievement.” European Journal of Psychology of Education 32: 39–59. https://
doi.org/10.1007/s10212-016-0298-5.
 STUDIES IN HIGHER EDUCATION 13

Díaz, Irene, Ana B. Bernardo, Maria Esteban, and Luis J. Rodríguez-Muñiz. 2020. “Variables Influencing University
Dropout: A Machine Learning-Based Study.” In The 11th International Conference on EUropean Transnational
Educational (ICEUTE 2020). Advances in Intelligent Systems and Computing, edited by Á Herrero, C. Cambra, D.
Urda, J. Sedano, H. Quintián, and E. Corchado, 93–104. Cham: Springer International Publishing.
Díaz Mújica, Alejandro, María V. Pérez Villalobos, Ana B. Antonio Cervero Fernández-Castañón, García González-Pienda,
and Antonio Julio. 2019. “Affective and Cognitive Variables Involved in Structural Prediction of University Dropout.”
Psicothema 31 (4): 429–36.
Diaz Lema, Melisa, Melvin Vooren, Marta Cannistrà, Chris van Klaveren, Tommaso Agasisti, and Ilja Cornelisz. 2023.
“Predicting Dropout in Higher Education Across Borders.” Studies in Higher Education, 1–16. https://doi.org/10.
1080/03075079.2023.2224818.
D. Molnár, Éva. 2013. Tudatos fejlődés. Az önszabályozott tanulás elmélete és gyakorlata. Budapest: Akadémiai Kiadó.
https://mersz.hu/?kdid=266.
D. Molnár, Éva, and Zita Gál. 2019. “Egyetemi Tanulmányaikat Megkezdő Hallgatók Tanulási Mintázata és Tanulói
Profilja.” Iskolakultúra 29 (1): 29–41. https://doi.org/10.14232/ISKKULT.2019.1.29.
Farruggia, Susan P., Cheon-woo Han, Lakeshia Watson, Thomas P. Moss, and Bette L. Bottoms. 2018. “Noncognitive
Factors and College Student Success.” Journal of College Student Retention: Research, Theory & Practice 20 (3): 308–
27. https://doi.org/10.1177/1521025116666539.
Funke, Joachim. 2001. “Dynamic Systems as Tools for Analysing Human Judgement.” Thinking & Reasoning 7 (1): 69–89.
https://doi.org/10.1080/13546780042000046.
Goswami, Usha. 1991. “Analogical Reasoning: What Develops? A Review of Research and Theory.” Child Development 62
(1): 1–22. https://doi.org/10.2307/1130701.
Greiff, Samuel, Sascha Wüstenberg, Benő Csapó, Andreas Demetriou, Jarkko Hautamäki, J. Arthur, C. Graesser, and
Romain Martin. 2014. “Domain-General Problem Solving Skills and Education in the 21st Century.” Educational
Research Review 13: 74–83. https://doi.org/10.1016/j.edurev.2014.10.002.
Greiff, Samuel, Sascha Wüstenberg, Gyöngyvér Molnár, Andreas Fischer, Joachim Funke, and Benő Csapó. 2013.
“Complex Problem Solving in Educational Contexts—Something Beyond g: Concept, Assessment, Measurement
Invariance, and Construct Validity.” Journal of Educational Psychology 105 (2): 364–79. https://doi.org/10.1037/
a0031856.
Hamers, Johan H. M., Els de Koning, and Klaas Sijtsma. 1998. “Inductive reasoning in third grade: Intervention promises
and constraints.” Contemporary Educational Psychology 23 (2): 132–48.
Hamoud, Alaa, Ali Salah Hashim, and Wid Akeel Awadh. 2018. “Predicting Student Performance in Higher Education
Institutions Using Decision Tree Analysis.” International Journal of Interactive Multimedia and Artificial Intelligence
5: 26–31. https://doi.org/10.9781/ijimai.2018.02.004.
Hódi, Ágnes, and Edit Tóth. 2019. “Elsőéves egyetemi hallgatók szövegértés fejlettsége és olvasási attitűdjei.”
Iskolakultúra 29 (1): 55–67.
Klauer, Karl Josef, and Gary D. Phye. 2008. “Inductive Reasoning: A Training Approach.” Review of Educational Research
78 (1): 85–123. https://doi.org/10.3102/0034654307313402.
Kuger, Susanne, Nina Jude, Eckhard Klieme, and David Kaplan. 2016. “An Introduction to the PISA 2015 Questionnaire
Field Trial: Study Design and Analysis Procedures.” In Assessing Contexts of Learning: An International Perspective,
edited by S. Kuger, E. Klieme, N. Jude, and D. Kaplan, 75–113. Cham: Springer.
Li, Kam Cheong, and Billy Tak-Ming Wong. 2019. “Factors Related to Student Persistence in Open Universities: Changes
Over the Years.” International Review of Research in Open and Distributed Learning 20 (4): 132–51.
Luszczynska, Aleksandra, Benicio Gutiérrez-Doña, and Ralf Schwarzer. 2005. “General Self-Efficacy in Various Domains of
Human Functioning: Evidence from Five Countries.” International Journal of Psychology 40 (2): 80–9. https://doi.org/
10.1080/00207590444000041.
Martos, Tamás, Balázs Jagodics, Judit Kőrössy, and Éva Szabó. 2021. “Psychological Resources, Dropout Risk and
Academic Performance in University Students–Pattern-Oriented Analysis and Prospective Study of Hungarian
Freshmen.” Current Psychology, 1–15.
Molnár, Gyöngyvér. 2011. “Playful Fostering of 6- to 8-Year-old Students’ Inductive Reasoning.” Thinking Skills and
Creativity 6 (2): 91–9. https://doi.org/10.1016/j.tsc.2011.05.002.
Molnár, Gyöngyvér, Saleh Ahmad Alrababah, and Samuel Greiff. 2022. “How we Explore, Interpret, and Solve Complex
Problems: A Cross-National Study of Problem-Solving Processes.” Heliyon 8 (1).
Molnár, Gyöngyvér, Samuel Greiff, and Benő Csapó. 2013. “Inductive Reasoning, Domain Specific and Complex Problem
Solving: Relations and Development.” Thinking Skills and Creativity 9: 35–45.
Molnár, Gyöngyvér, Samuel Greiff, Sascha Wüstenberg, and Andreas Fischer. 2017. “Empirical study of computer based
assessment of domain-general dynamic problem solving skills.” In The Nature of Problem Solving, edited by B. Csapó,
and J. Funke, 123–43. Paris: OECD.
Molnár, Gyöngyvér, Ágnes Hódi, Éva D. Molnár, Zoltán Nagy, and Benő Csapó. 2021. “Assessment of First-Year University
Students: Facilitating an Effective Transition Into Higher Education.” In Új kutatások a neveléstudományokban 2020,
edited by Á Engler, and V. Bocsi, 11–26. Debrecen.
14 G. MOLNÁR AND Á. KOCSIS

Molontay, Roland, and Marcell Nagy. 2023. “How to Improve the Predictive Validity of a Composite Admission Score? A
Case Study from Hungary.” Assessment & Evaluation in Higher Education 48 (4): 419–37. https://doi.org/10.1080/
02602938.2022.2093835.
Mousa, Mojahed, and Gyöngyvér Molnár. 2020. “Computer-based Training in Math Improves Inductive Reasoning of 9-
to 11-Year-old Children.” Thinking Skills and Creativity 37: 100687. https://doi.org/10.1016/j.tsc.2020.100687.
Musso, Mariel F., Carlos Felipe Rodríguez Hernández, and Eduardo C. Cascallar. 2020. “Predicting key Educational
Outcomes in Academic Trajectories: A Machine-Learning Approach.” Higher Education 80: 875–94. https://doi.org/
10.1007/s10734-020-00520-7.
Muthén, Linda K., and Bengt O. Muthén. 2012. Mplus User’s Guide. 7th ed. Los Angeles, CA: Muthén and Muthén.
Naaman, Hind. 2021. “The Academic Dropout Wheel Analyzing the Antecedents of Higher Education Dropout in
Education Studies.” The European Educational Researcher 4 (2): 133–53. https://doi.org/10.31757/euer.421.
Ndoye, Abdou, Shawn Clarke, and Cori Henderson. 2020. “Predicting College Students’ Academic Success.” Journal of
Student Success and Retention 6 (1): 36–61.
OECD. 2019. Education at a Glance 2019: OECD Indicators. Paris: OECD.
OECD. 2022. Education at a Glance 2022: OECD Indicators. Paris: OECD.
OECD. 2010. Education at a glance 2010: OECD indicators. Paris: OECD.
OECD/European Union. 2017. “Overview of the Hungarian Higher Education System.” In Supporting Entrepreneurship
and Innovation in Higher Education in Hungary. Paris/European Union, Brussels: OECD Publishing.
Olivier, Elizabeth, Isabelle Archambault, Mikaël De Clercq, and Benoit Galand. 2019. “Student Self-Efficacy, Classroom
Engagement, and Academic Achievement: Comparing Three Theoretical Frameworks.” Journal of Youth and
Adolescence 48 (2): 326–40.
Panadero, Ernesto. 2017. “A Review of Self-Regulated Learning: Six Models and Four Directions for Research.” Frontiers in
Psychology 8: 422. https://doi.org/10.3389/fpsyg.2017.00422.
Pastén, Laura Espinoza. 2021. “Metacognitive, Critical and Creative Thinking in Educative Contexts: Conceptualization
and Didactic Suggestions.” Psicologia Escolar e Educacional 25 (1).
Pásztor, Attila, Sirkku Kupiainen, Risto Hotulainen, Gyöngyvér Molnár, and Benő Csapó. 2018. “Comparing Finnish and
Hungarian Fourth Grade Students’ Inductive Reasoning Skills.” Paper presented at 9th EARLI SIG 1 Conference,
Helsinki, Finland, August 2018.
Pintrich, Paul R. 1999. “The Role of Motivation in Promoting and Sustaining Self-Regulated Learning.” International
Journal of Educational Research 31 (6): 459–70. https://doi.org/10.1016/S0883-0355(99)00015-4.
Pintrich, Paul R. 2000. “The Role of Goal Orientation in Self-Regulated Learning.” In Handbook of Self-Regulation, edited
by M. Boekaerts, P. R. Pintrich, and M. Zeidner, 451–502. Orlando, FL: Academic Press.
Pinxten, Maarten, Bieke D. Fraine, Wim Van Den Noortgate, Jan Van Den Damme, Tinneke Boonen, and Gudrun Vanlaar.
2015. “‘I Choose so I Am’: A Logistic Analysis of Major Selection in University and Successful Completion of the First
Year.” Studies in Higher Education 40 (10): 1919–46. https://doi.org/10.1080/03075079.2014.914904.
Pirttimaa, Raija, Marjatta Takala, and Tarja Ladonlahti. 2015. “Students in Higher Education with Reading and Writing
Difficulties.” Education Inquiry 6 (1): 24277. https://doi.org/10.3402/edui.v6.24277.
Ribeiro, Luísa, Pedro Rosário, José C. Núñez, Martha Gaeta, and Sonia Fuentes. 2019. “First-year Students Background
and Academic Achievement: The Mediating Role of Student Engagement.” Frontiers in Psychology 10: 2669.
https://doi.org/10.3389/fpsyg.2019.02669.
Richardson, Michelle, Charles Abraham, and Rod Bond. 2012. “Psychological Correlates of University Students’ Academic
Performance: A Systematic Review and Meta-Analysis.” Psychological Bulletin 138 (2): 353–87. https://doi.org/10.
1037/a0026838.
Roberts, Steven. 2011. “Traditional practice for non-traditional students? Examining the role of pedagogy in higher edu-
cation retention.” Journal of Further and Higher Education 35 (2): 183–199.
Rode Larsen, Malene, Hanna Bjørnøy Sommersel, and Michael Søgaard Larsen. 2013. Evidence on Dropout Phenomena at
Universities: Malene Rode Larsen, Hanna Bjørnøy Sommersel, Michael Søgaard Larsen. Copenhagen: Danish
Clearinghouse for Educational Research. https://edudoc.ch/record/115243/files/Dropout_from_universities_brief_
version.pdf.
Rodríguez-Hernández, Carlos Felipe, Eduardo Cascallar, and Eva Kyndt. 2020. “Socio-economic Status and Academic
Performance in Higher Education: A Systematic Review.” Educational Research Review 29: 100305. https://doi.org/
10.1016/j.edurev.2019.100305.
Rodríguez-Hernández, Carlos Felipe, Mariel Musso, Eva Kyndt, and Eduardo Cascallar. 2021. “Artificial Neural Networks in
Academic Performance Prediction: Systematic Implementation and Predictor Evaluation.” Computers and Education:
Artificial Intelligence 2: 100018. https://doi.org/10.1016/j.caeai.2021.100018.
Rump, Markus, Wiebke Esdar, and Elke Wild. 2017. “Individual Differences in the Effects of Academic Motivation on
Higher Education Students’ Intention to Drop Out.” European Journal of Higher Education 7 (4): 341–55. https://
doi.org/10.1080/21568235.2017.1357481.
Schneider, Michael, and Franzis Preckel. 2017. “Variables Associated with Achievement in Higher Education: A
Systematic Review of Meta-Analyses.” Psychological Bulletin 143 (6): 565. https://doi.org/10.1037/bul0000098.
 STUDIES IN HIGHER EDUCATION 15

Van Herpen, Sanne G. A., Marieke Meeuwisse, Adriaan W. H. Hofman, and Sabine E. Severiens. 2020. “A Head Start in
Higher Education: The Effect of a Transition Intervention on Interaction, Sense of Belonging, and Academic
Performance.” Studies in Higher Education 45 (4): 862–77. https://doi.org/10.1080/03075079.2019.1572088.
Van Rooij, Els C.M., Ellen P. W. A. Jansen, and Wim J. C. M. van de Grift. 2018. “First-year University Students’ Academic
Success: The Importance of Academic Adjustment.” European Journal of Psychology of Education 33: 749–67. https://
doi.org/10.1007/s10212-017-0347-8.
Vanthournout, Gert, David Gijbels, Liesje Coertjens, Vincent Donche, and Peter Van Petegem. 2012. “Students’
Persistence and Academic Success in a First-Year Professional Bachelor Program: The Influence of Students’
Learning Strategies and Academic Motivation.” Education Research International 12.
Weinstein, Claire Ellen, Jenefer Husman, and Douglas R. Dierking. 2000. “Self-Regulation Interventions with a Focus on
Learning Strategies.” In Handbook of Self-Regulation, edited by M. Boekaerts, M. Zeidner, and P. R. Pintrich, 727–47.
Academic Press.
Westrick, Paul A., Frank L. Schmidt, Huy Le, Steven B. Robbins, and Justine M. R. Radunzel. 2021. “The Road to Retention
Passes Through First Year Academic Performance: A Meta-Analytic Path Analysis of Academic Performance and
Persistence.” Educational Assessment 26 (1): 35–51. https://doi.org/10.1080/10627197.2020.1848423.
Winne, Philip H., and Nancy E. Perry. 2000. “Measuring Self-Regulated Learning.” In Handbook of Self-Regulation, edited
by M. Boekaerts, M. Zeidner, and P. R. Pintrich, 531–66. Academic Press.
Wu, Hao, and Gyöngyvér Molnár. 2018. “Interactive Problem Solving: Assessment and Relations to Combinatorial and
Inductive Reasoning.” Journal of Psychological & Educational Research 26 (1).
Wu, Hao, Andi Rahmat Saleh, and Gyöngyvér Molnár. 2022. “Inductive and Combinatorial Reasoning in International
Educational Context: Assessment, Measurement Invariance, and Latent Mean Differences.” Asia Pacific Education
Review 23 (2): 297–310. https://doi.org/10.1007/s12564-022-09750-z.
York, Travis T., Charles Gibson, and Susan Rankin. 2015. “Defining and Measuring Academic Success.” Practical
Assessment, Research, and Evaluation 20 (1): 5.
Yu, Ching-Yun. 2002. Evaluating Cutoff Criteria of Model fit Indices for Latent Variable Models with Binary and Continuous
Outcomes. Los Angeles.: University of California.
Zlatkin-Troitschanskaia, Olga, Richard J. Shavelson, and Christiane Kuhn. 2015. “The International State of Research on
Measurement of Competency in Higher Education.” Studies in Higher Education 40 (3): 393–411. https://doi.org/10.
1080/03075079.2015.1004241.

Annex 1. Demographic variables of participating and non-participating students.

Participating (N = 1681) Non-participating (1732)

Min. Max. Mean SD Min. Max. Mean SD t p
Sex 1 2 1.51 0.5 1 2 1.56 0.5 −7.4 <.001
Age 18 47 19.86 2.11 18 53 20.71 2.99 3.05 <.01
Entry score 280 493 381.6 50.27 280 488 369.0 48.40 9.62 <.001