Dropping out of higher education: Analysis of variables that characterise students who interrupt their studies

Abstract

One of the persistent problems faced by the university education system is the dropout rate. The main aim of this research was to identify the profile characteristics of those students who drop out of their studies, seeking in-depth knowledge of the reality behind the issue. The responses to a questionnaire of 149,837 students from three Spanish universities (La Laguna, Zaragoza and Huelva) who had dropped out of their undergraduate studies were analysed. The outcomes enabled us to identify a number of features associated with the likelihood of dropping out of university studies. Specifically, it was found that the university of study, sex, age, study branch, entry qualification, scholarship or grant, nationality and job are predictors of dropout. The results obtained have an important transfer value, with a view to implementing actions for the adaptation of students and to avoid university dropout.

Full text

Dropping out of higher education: Analysis of variables that characterise
students who interrupt their studies
María Olga Gonz´
alez-Morales
a
, David L´
opez-Aguilar
b,*
, Pedro Ricardo ´
Alvarez-P´
erez
b
,
Pedro Antonio Toledo-Delgado
c
a
Department of Applied Economics and Quantitative Methods, University of La Laguna, Avenida Universidad, s/n. 38206, San Crist´
obal de La Laguna, Canary Islands,
Spain
b
Department of Didactic and Educational Research, University of La Laguna, Avenida Universidad, s/n. 38206, San Crist´
obal de La Laguna, Canary Islands, Spain
c
Department of Computer and Systems Engineering, University of La Laguna, Avenida Universidad, s/n. 38206, San Crist´
obal de La Laguna, Canary Islands, Spain
ARTICLE INFO
Keywords:
Higher education
University students
Academic dropout
Risk factors
ABSTRACT
One of the persistent problems faced by the university education system is the dropout rate. The main aim of this
research was to identify the prole characteristics of those students who drop out of their studies, seeking in-
depth knowledge of the reality behind the issue. The responses to a questionnaire of 149,837 students from
three Spanish universities (La Laguna, Zaragoza and Huelva) who had dropped out of their undergraduate studies
were analysed. The outcomes enabled us to identify a number of features associated with the likelihood of
dropping out of university studies. Specically, it was found that the university of study, sex, age, study branch,
entry qualication, scholarship or grant, nationality and job are predictors of dropout. The results obtained have
an important transfer value, with a view to implementing actions for the adaptation of students and to avoid
university dropout.
1. Introduction
When analysing different parameters for assessing the quality of
higher education, the number of students who drop out of education and
the causes behind this issue are aspects that receive a lot of attention
(Gallego et al., 2021). Reducing university dropout rates and improving
student retention has therefore become one of the major challenges
facing university institutions each year (De la Cruz-Campos et al., 2023;
Lizarte & Gij´
on, 2022; Matta et al., 2017; Pusztai et al., 2022). Rather
than competing with other institutions, the problem that universities
must solve concerns nding a way to avoid losing a signicant per-
centage of students who give up their studies before completion.
Although the international literature provides various conceptualisa-
tions (B¨
aulke et al., 2021), in this study, dropout is understood as the
situation of students who, after enrolling for the rst time in a degree
programme, do not renew their enrolment for the following two years,
leaving the university denitively (Kehm et al., 2019).
As stated by Parra-S´
anchez et al. (2023), dropping out of university
should certainly be seen as a global phenomenon, occurring worldwide.
This is conrmed in research such as that by Freitas et al. (2022, p.8) or
´
Alvarez (2021), who noted that academic “dropout is a very relevant
phenomenon” in different continents, the consequences of which are felt
not only in academic contexts, but also in society at large (Ahn & Davis,
2020). According to the OECD (2016), countries such as Hungary, New
Zealand and the United States have dropout rates of around 50 %,
compared to other countries such as Australia, Denmark and Japan,
where dropout rates are around 20 %. According to Acevedo (2021),
dropout rates in higher education institutions in Latin America are also
very high. In the European scope, Spain has fairly high university
dropout rates. According to the Ministry of Universities (2022), the
percentage of university students who drop out is over 15 %. Likewise,
according to CRUE (2022), dropout rates in on-site public universities
are above 20 % and exceed 40 % in off-site public universities. In the
latest report published by the Spanish Ministry of Universities
(Fern´
andez-Mellizo, 2022), the dropout rate from undergraduate studies
in on-site universities was 13 % of the student body. These gures give
an idea of the impact exerted by this problem. In terms of dropout rates
according to branches of knowledge (CRUE, 2022), the most noteworthy
* Corresponding author.
E-mail addresses: olgonzal@ull.edu.es (M.O. Gonz´
alez-Morales), dlopez@ull.edu.es (D. L´
opez-Aguilar), palvarez@ull.edu.es (P.R. ´
Alvarez-P´
erez), petode@ull.edu.
es (P.A. Toledo-Delgado).
Contents lists available at ScienceDirect
Acta Psychologica
journal homepage: www.elsevier.com/locate/actpsy
https://doi.org/10.1016/j.actpsy.2024.104669
Received 29 June 2024; Received in revised form 12 November 2024; Accepted 11 December 2024
Acta Psychologica 252 (2025) 104669
Available online 17 December 2024
0001-6918/© 2024 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC license (
http://creativecommons.org/licenses/by-
nc/4.0/ ).

are Engineering and Humanities (27.19 %), Arts and Humanities (24.05
%), Science (19.8 %) and Social and Legal Sciences (18.40 %). However,
the trend of low dropout rates in Health Sciences (12.01 %) continues.
Characterising university dropout is a complex problem. Studies
carried out so far have shown the relationship of dropout with different
types of variables (Aina et al., 2022; Constante-Amores et al., 2021;
Pe˜
na-V´
azquez et al., 2023). Low academic performance (Casanova et al.,
2018), economic problems (Bertola, 2023), disadvantaged social con-
texts (Calamet, 2020), lack of integration at university (Esteban et al.,
2017), stress (Eicher et al., 2014), low academic motivation (Thuy et al.,
2017), relationships with faculty and peers (Bernardo et al., 2016), low
academic engagement (´
Alvarez et al., 2021), poor self-regulation of
learning (Merino-Tejedor et al., 2016), procrastination, and dissatis-
faction with studies (Lindner et al., 2023; Merino-Tejedor et al., 2016),
emotional problems (P´
erez et al., 2022) or low satisfaction with the
teaching received (´
Alvarez & L´
opez-Aguilar, 2020). These are variables
that are positively related to dropout and explain why a high percentage
of students give up their studies before completing them.
We are therefore faced with a multifactorial phenomenon, in which
variables of a very diverse nature are related, which in most cases
combine and lead to the interruption of studies (Schneider & Preckel,
2017). Research by Rodríguez-Pineda and Zamora-Araya (2021) shows
the complex structure of factors involved in dropout. Through an
exploratory factor analysis, they concluded that dropout was related to
academic, motivational, economic, family and vocational variables. The
study by ˇ
Sabi´
c and Puzi´
c (2022) also highlights that the risk of dropping
out is related to several factors (social background, income, academic
and institutional characteristics). In the study by Lindner et al. (2023), a
longitudinal study was carried out to positively conrm the reciprocal
relationships between procrastination, satisfaction with studies and the
propensity to withdraw from their studies among university students.
According to the authors, procrastination has negative consequences for
university students (it is positively related to dissatisfaction and, in turn,
dissatisfaction is related to intention to drop out).
If we analyse some of the demographic, academic and economic
characteristics of students in depth, several studies (Benítez et al., 2021;
Cuji et al., 2023; Díaz et al., 2021; Zamora et al., 2023) indicate that
variables such as gender, age, nationality, the type of studies students
undertake, academic performance, the way they nance their studies or
the combination of studies and work are factors to be considered when
analysing the probability of students dropping out of university.
Regarding sex and age, there are no conclusive studies, as the nd-
ings are either one way or the other or no signicant variables have been
found. In this sense, Casanova et al. (2018) noted that women are more
likely to give up their university studies, whereas Gairín et al. (2014)
indicated that it is men who are more likely to drop out. As for age,
C´
espedes-L´
opez et al. (2021) argue that the older the student, the greater
the likelihood of dropout. Nevertheless, Esteban et al. (2017) failed to
determine its inuence on dropout rates. Studies analysing the possible
inuence of having a scholarship/grant or not also reach conicting
results. In any case, authors such as Fern´
andez-Martín et al. (2019) or
Constante-Amores et al. (2021) concluded that age is a variable with a
high incidence in dropout. Regarding employment status, research is
more decisive and a strong association has been found between student
employment and school dropout (an increased likelihood of withdrawal
from studies among students who are employed) (Hern´
andez & Vargas,
2016; Freixa et al., 2018; Tuero et al., 2018).
Although academic dropout is not a new issue, the fact is that year
after year statistics continue to be reproduced that alert us to the seri-
ousness of the problem. This gives rise to continuous calls for studies to
help clarify the nature of this phenomenon and to identify variables that
can help predict it (Huo et al., 2023; Li & Carroll, 2020). In this way,
policies and programmes could be put in place to contribute to the
integration and retention of students entering university, so that they are
able to graduate upon completion of their studies (De Silva et al., 2022;
Piepenburg & Beckmann, 2021). This would provide people with the
skills they need to enter the workplace. As P´
erez and Ald´
as (2019) point
out, in Spain around 45.5 % of young people go on to higher education,
but only 32.9 % manage to graduate. And it cannot be overlooked that
an important objective set by the European Union is that 40 % of the
population between 30 and 35 years of age should achieve a qualica-
tion for employment (European Commission et al., 2015). For this
reason, it is just as important for young people to choose university
studies as it is for them to complete the university education they choose
to pursue. To this end, it is necessary to analyse the variables with which
dropout is related, in order to implement adaptation strategies before
dropout occurs. In relation to this problem, the main purpose of this
research was to identify which variables (academic, personal and
contextual) predict university dropout.
2. Methodology
2.1. Hypothesis
Setting out from a review of previous research that has explored this
problem in depth, and with the aim of determining which factors and
personal traits may inuence the likelihood of dropping out of higher
education, the following hypotheses were put forward:
H1. The likelihood of a student dropping out of a university degree
increases in students (demographic characteristics):
H1.1. Men
H1.2. Foreigners
H1.3. Aged below 25 years.
H2. The likelihood of a student dropping out of a university degree
increases in students (academic characteristics):
H2.1. from the Social and Legal Sciences branch
H2.2. having gained access to university with low entry grades
H3. The likelihood of a student dropping out of a university degree
increases in students (economic characteristics):
H3.1. without a scholarship or grant
H3.2. that are working
2.2. Participants
The study population was the 150,018 university students who
enrolled in the rst year of a degree course (period 2010–2018) at the
Universities of Huelva, La Laguna and Zaragoza (Spain). The survey was
carried out at the end of each academic year. These universities
participated in the research project entitled “Analysis of the explanatory
factors of university dropout and strategic actions for its improvement
and prevention” (reference: “Analysis of the explanatory factors of
university dropout and strategic actions for its improvement and pre-
vention”): PID2020-114849RB-I00) funded by the Ministry of Science
and Innovation of the Government of Spain. After the process of clearing
the database containing the information used for the analyses, a total of
149,837 students took part in the study. To clear the database, the
variables and cases were taken into account. Those variables with a high
number of non-response cases were removed from the analysis. As for
the cases, those containing incorrect answers were deleted (no infor-
mation related to the variable). Subsequently, the remaining variables
were examined and those in which all cases had responded were selected
and different groupings were made according to whether or not the
variable was signicant with respect to the variable to be used in the
analyses as the dependent variable (whether or not the student dropped
out of the degree programme). Of the total study participants, 44.7 % (n
=67,046) were male and 55.3 % female (n =82,791). Some 86.4 % (n
=129,439) were <25 years of age and 13.6 % (n =20,398) were older.
M.O. Gonz´
alez-Morales et al.
Acta Psychologica 252 (2025) 104669
2

The characteristics of the participating sample are presented in the table
below (Table 1).
2.3. Study variables
The variables that were subjected to analysis are presented in
Table 2. Specically, this table shows the type of variables (dependent/
independent), their name, the categorisation made and the percentage
obtained.
These variables were gathered from the different data centres of the
universities participating in the study. Although it is true that each
institution had more information on the student body, in order to carry
out a unied study, it was decided to use only the variables that were
common to the different universities. Prior to accessing the data, the
University of La Laguna, which was the institution from which the
research project was managed and coordinated, asked the Research
Ethics and Animal Welfare Committee (CEIBA) for a report conrming
that the work met the ethical requirements for its development. The
committee issued a favourable resolution (reference: CEIBA2021–3079)
for the study to be carried out. Likewise, the data protection ofcers of
the universities participating in the project established the procedures to
be followed in order to access the information under study. In particular,
the researchers of the research project signed a condentiality agree-
ment in which they were responsible for using the information provided
by the data centres solely and exclusively for the purposes of this study.
With these preliminary steps in place, the researchers from the
different universities participating in the project contacted their
respective data centres, who provided the necessary information about
the variables listed in Table 2. Some of the variables used in the analysis
were not explicitly recorded in the original data repositories, and were
calculated using the following procedures:
The dependent variable “degree course dropout” was calculated from
the ofcial denition of the Ministry of Universities of the Spanish
Government, which indicates that a student has dropped out when
“without having graduated from that degree they have not enrolled in it
for two consecutive years”.
The databases of the participating universities had information on
students in different administrative situations:
- Students who graduated in any degree.
- Students enrolled in each academic year.
- Students who transferred their le to another institution.
Those students who were listed as graduates were assigned the value
“No” to the “degree course dropout” variable. Of the remaining students,
the following were ruled out of the study:
- Those students who were enrolled in one of the last two academic
years, as they are students with studies ongoing and it is not known
whether they dropped out or not.
- Students who had nally transferred to another university, as it was
not known whether or not they completed their studies at the new
university.
The remaining students, having met the criteria of having enrolled in
a university degree, not having graduated and not having requested a
transfer of their transcript to another institution, were categorised with
the value “Yes” in the “degree course dropout” variable.
As for the independent variables, the original coding is directly
convertible to the values described below, with the exception of
“scholarship/grant holder” and “work activity”. To determine the
“scholarship/grant holder” eld, an assignment was made according to
the type of grant they had, in such a way that students who had a grant,
regardless of the type, were assigned the value “Yes”. Similarly, for the
“work activity” variable, the “Yes” coding was carried out in those cases
in which the students engaged in a professional activity, whatever type
of activity it might be.
2.4. Ethical issues
A favourable report was received from the Research Ethics and An-
imal Welfare Committee (CEIBA). In collaboration with the Data Pro-
tection Delegate of the participating institutions, an informed consent
Table 1
Participating sample characteristics.
Sex Men: 44.7 % (n =67,046)
Women: 55.3 % (n =82,791)
Age <25 years: 86.4 % (n =129,439)
25 years or more: 13.6 % (n =20,398)
Knowledge branch Arts and Humanities: 9.6 % (n =14,414)
Science: 7.1 % (n =10,705)
Health Sciences: 14.7 % (n =22,033)
Social and Legal Sciences: 48.7 % (n =72,955)
Engineering and Architecture: 19.8 % (n =29,730)
University University of Huelva: 16.9 % (n =25,383)
University of La Laguna: 34.3 % (n =51,440)
University of Zaragoza: 48.7 % (n =73,014)
Table 2
Variables used in the analysis.
Type Variable Category Percentage
Dependent Degree course dropout 0 =No 82.5 (n =
123,577)
1 =Yes 17.5 (n =
26,260)
Independent
University
1 =University of
Huelva
16.9 (n =
25,383)
2 =University of
Zaragoza
48.7 (n =
73,014)
3 =University of La
Laguna
34.3 (n =
51,440)
Sex
1 =Male 44.7 (n =
67,046)
2 =Female 55.3 (n =
82,791)
Age
1 =<25 years 86.4 (n =
129,439)
2 =25 years or more 13.6 (n =
20,398)
Study branch
1 =Engineering and
Architecture
19.8 (n =
29,730)
2 =Health Sciences 14.7 (n =
22,033)
3 =Science 7.1 (n =
10,705)
4 =Arts and
Humanities
9.6 (n =
14,414)
5 =Social and Legal
Sciences
48.7 (n =
72,955)
University entrance
qualication
1 =(5–7) 39.1 (n =
58,530)
2 =(7–9) 26.9 (n =
40,344)
3 =(9–10) 10.8 (n =
16,220)
4 ≥10 23.2 (n =
34,743)
Has a scholarship/
grant
0 =No 64.2 (n =
96,202)
1 =Yes 35.8 (n =
53,635)
Nationality
1 =Foreign 28 (n =4237)
2 =Spanish 97.2 (n =
145,600)
Working activity 1 =No 96.3 (n =
144,332)
2 =Yes 3.7 (n =5505)
M.O. Gonz´
alez-Morales et al.
Acta Psychologica 252 (2025) 104669
3

and condentiality commitment document was available for use in the
research process. These documents were employed as a preliminary step
prior to engaging with the participants. Their purpose was to commu-
nicate the study's objectives and to ensure the condentiality of the
information collected.
2.5. Analysis techniques
For the analysis of the characteristics of the student who dropped out
of university, it was necessary to previously select those variables that
presented signicant differences between the groups analysed.
To assess independence among the variables, contingency tables
were drawn up and the Chi-square test applied. In this type of test, the
null hypothesis (H0) posits that the two variables are independent, while
the alternative hypothesis (H1) suggests that there is some level of as-
sociation or relationship between the variables. When looking for sig-
nicant differences among the variables, depending on whether or not
the student drops out of university studies, the probability is expected to
be <0.05, so that the observed distribution does not behave like the
expected distribution and, thus, the variables with signicant and,
therefore, independent differences can be used in the subsequent anal-
ysis. Table 3 summarises the percentages of each variable, depending on
whether or not the student had dropped out of school, and the results of
the Chi-square test, which were signicant for all variables.
A Binary Logistic Regression (BLR) analysis was then applied. BLR is
used to test hypotheses or causal relationships when the dependent
variable is a binary/dichotomous variable, i.e. it has only two cate-
gories. BLR analysis is based on principles such as odds ratios and
probabilities. Independent variables try to predict the probability of
some event occurring over the likelihood of it not occurring.
The aim of this work is to look for those characteristics of students
that may inuence them to drop out of university studies. The
dependent variable, degree course dropout, has two categories: 0 =No
and 1 =Yes.
The model proposed in this paper is:
Y =pro{yes}=1
1+e−z where Z is the following linear combination
Z=b0+b1X1+b2X2+b3X3+b4X4+b5X5+b6X6+b7X7+b8X8+
ε
Y =probability of occurrence of the event, dependent variable
X1, X2, X3, X4, X5, X6, X7, X8 =scores of the independent variables
b0 =constant
b1,b2,b3,b4,b5,b6,b7,b8 =estimated regression coefcient estimates
that report how much the probability of occurrence of Z varies in the
face of a unit change of each independent variable, all other variables
remaining constant.
ε
=estimation error.
To corroborate that the results of the BLR are feasible, it is necessary
to analyse the goodness of t of the model and the relationship of the
independent variables on the dependent variable. The Enter method,
which uses all the variables in a single step, was applied.
On the goodness of t of the model, it is necessary to analyse:
1. Chi-square signicance of the model in the omnibus test. Signi-
cance <0.05 shows that the model helps explain the event, i.e. the
independent variables explain the behaviour of the dependent vari-
able. In this study, the outcome was signicant (0.000).
2. “Cox and Snell's R-squared and Nagelkerke's R-squared. These indi-
cate the part of the variance of the dependent variable explained by
the model. The part of the dependent variable explained by the
model is interpreted as ranging between the Cox and Snell R-squared
and the Nagelkerke R-squared” (Marco-Franco, 2022, p.93). The
higher the R-squared, the more explanatory the model is. The results
range from 16.7 % (0.167 Cox and Snell R-squared) to 29.7 % (0.297
Nagelkerke R-squared).
3. “Overall percentage correctly classied, i.e. the number of cases that
the model is able to predict correctly. Based on the regression
equation and the observed data, a prediction of the value of the
dependent variable (predicted value) is made, which is compared
with the observed value; if it is correct, the case is correctly classi-
ed” (Medel-Ramírez & Medel-L´
opez, 2018, p.41). The closer the
predicted value matches the observed value, the more explanatory
the model is, i.e. the more the independent variables are good pre-
dictors of the dependent variable. In this case, 82.2 % were correctly
classied.
Regarding the relationship of the independent variables with the
dependent variable, the following should be considered:
1. The individual signicance of each of the parameters (B). If it
is<0.05, that independent variable explains the dependent variable.
2. The sign of B shows the direction of the relationship.
3. As we did in other work (Pe˜
na-V´
azquez et al., 2023, p.301), for
interpretation purposes, Exp (B), exponential of B, represents the
odds ratio and indicates the strength of the relationship (the further
away from 1, the stronger the relationship). “When Exp (B) is >1, it
indicates that an increase in the independent variable increases the
odds of the event (dependent variable) occurring. When Exp (B) is
<1, it indicates that an increase in the independent variable reduces
the odds of the event (dependent variable) occurring”. To compare
the exponentials of B with each other, those that are <1 must be
transformed into their inverse or reciprocal, i.e., we must divide 1 by
the value of Exp (B).
Table 3
Student features according to whether or not they drop out of the degree pro-
gramme, and Chi-square results.
Variable Category Drops out Chi-square
No Yes
University
University of Huelva 16.4 19.7
Х
2
(2) =
650.055
p ≤.000
University of
Zaragoza 50.2 41.6
University of La
Laguna 33.4 38.7
Sex
Male 42.2 56.7 Х
2
(1) =
1836.289
p ≤.000
Female 57.8 43.3
Age
<25 years 89.7 70.9 Х
2
(1) =
6475.311
p ≤.000
25 years or more 10.3 29.1
Study branch
Engineering and
Architecture 18.6 25.6
Х
2
(4) =
3095.746
p ≤.000
Health Sciences 16.7 5.1
Science 7.4 6.1
Arts and Humanities 8.8 13.6
Social and Legal
Sciences 48.5 49.6
University entrance
qualication
(5–7) 34.2 62.0 Х
2
(3) =
9571.894
p ≤.000
(7–9) 26.9 27.2
(9–10) 11.9 5.9
>10 27.1 4.9
Has a scholarship/grant
No 63.4 67.8 Х
2
(1) =
182.815
p ≤.000
Yes 36.6 32.2
Nationality
Foreign 2.6 3.7 Х
2
(1) =
91–567
p ≤.000
Spanish 97.4 96.3
Working activity
No 97.0 93.1 Х
2
(1) =
927.633
p ≤.000
Yes 3.0 6.9
M.O. Gonz´
alez-Morales et al.
Acta Psychologica 252 (2025) 104669
4

3. Results
Table 4 shows the results of the BLR concerning the variables within
the equation.
To achieve a model whose variables have greater explanatory power
for the dependent variable, the backward stepwise method was used,
which introduces all the predictor variables at the start of the analysis.
At each step, the least signicant term is removed from the model.
Thereafter, the algorithm alternates between forward entry of out-of-
model terms and backward elimination of stepwise terms from the
model. This continues until there are no terms left that meet the inclu-
sion or deletion criteria. However, in this analysis, all variables were
signicant in the rst step. The results indicate the following:
1) University. The negative value of B and the result of Exp (B) indicate
that a student is 1.1669 times less likely to drop out at the University
of Huelva than at the University of La Laguna and 1.4451 times less
likely at the University of Zaragoza.
2) Sex. The positive value of B and the result of Exp (B) indicate that a
male student is 1.548 times more likely to drop out of university than
a female student.
3) Age. The negative value of B and the result of Exp (B) indicate that a
student under 25 is 2.3259 times less likely to drop out than an older
student.
4) Knowledge branch. The negative value of B and the result of Exp (B)
indicate that a student from the Health Sciences branch of knowledge
is 2.2272 times less likely to drop out of university than a student
from Social and Legal Sciences. Meanwhile, students from Sciences
and Arts and Humanities presented a higher probability (1.107 and
1.442 respectively). The Engineering and Architecture branch was
not signicant.
5) University entry qualication. The positive value of B and the result
of Exp (B) indicate that the lower the university entry grade, the
more likely the student is to drop out. Students with a grade between
5 and 7 are 6.151 times more likely to drop out than a student with a
grade higher than 10. Students with a grade higher than 7 and lower
than 9 are 4.304 times more likely to drop out and those with grades
between 9 and 10 are 2.318 times more likely to drop out.
6) Scholarship/grant. The positive value of B and the result of Exp (B)
indicate that a student without a scholarship or grant is 1.130 times
more likely to drop out of university than a student awarded a
scholarship/grant.
7) Nationality. The positive value of B and the result of Exp (B) indicate
that a foreign student is 1.216 times more likely to drop out of uni-
versity than a Spanish student.
8) Working activity. The negative value of B and the result of Exp (B)
show that a student who does not work is 1.4728 times less likely to
drop out of school than a student who works.
4. Discussion and conclusions
The main objective of this research was to identify the characteristics
from which the prole of students who interrupt higher education can
be obtained and determine the probability of dropping out. To this end,
the study used variables related to the demographic, academic and
economic features of the student body. In terms of the hypotheses pro-
posed, it should be noted that, in general, all the variables studied were
signicant and inuential in the probability of dropping out of the de-
gree programme. However, despite being signicant, a more precise and
detailed analysis of the hypotheses and the results obtained revealed
that the rst and second hypotheses were partially fullled, while the
third hypothesis was fully met.
The demographic characteristics analysis indicated that the probability
of a student dropping out of a university degree increases for male
students and students from other (foreign) countries. Regarding the
gender variable and the “male students” characteristic, we found
agreement with the results obtained by Lindner et al. (2023). However,
it is not the under 25 s who drop out most, but the older ones, so that
H1.3 is not fullled. Although there is a coincidence of results with those
reported by Gairín et al. (2014) and C´
espedes-L´
opez et al. (2021), in the
case of age there is a discrepancy with the ndings of Lindner et al.
(2023), who state that “age” is not related to dropping out, although
years of schooling is related to “age”.
Regarding academic characteristics, students from the Social and
Legal Sciences branch are more likely to drop out, but only with respect
to those from Health Sciences (not compared to the rest of the branches).
These results do not coincide with those reported by Constante-Amores
et al. (2021), who argued that the Arts and Humanities and Social and
Legal Sciences branches are the ones with the highest dropout rates.
Nevertheless, they are in line with the results achieved by Pe˜
na-V´
azquez
et al. (2023), who showed that the students most likely to drop out were
Table 4
BLR analysis results. Variables in the equation.
B E.T. Wald gl Sig. Exp(B) I.C. 95,0 % EXP(B)
Inferior Superior Inferior Superior Inferior Superior Higher Lower
Step 1 (a)
University
440,817 2 ,000
Huelva University −,155 ,020 57,458 1 ,000 ,857 ,823 ,892
Zaragoza University −,367 ,018 440,680 1 ,000 ,692 ,669 ,717
Male ,437 ,015 808,968 1 ,000 1548 1502 1595
Under 25 years old −,845 ,018 2101,991 1 ,000 ,430 ,414 ,445
Branch 1065,852 4 ,000
Engineering and Architecture ,009 ,019 ,206 1 ,650 1009 ,972 1047
Health Sciences −,801 ,031 657,519 1 ,000 ,449 ,422 ,477
Science ,102 ,031 11,097 1 ,001 1107 1043 1176
Arts and Humanities ,366 ,023 257,429 1 ,000 1442 1379 1508
Entry grade 4088,536 3 ,000
[5–7] 1817 ,031 3421,873 1 ,000 6151 5787 6536
(7–9) 1460 ,032 2081,694 1 ,000 4304 4043 4583
[9–10] ,841 ,039 455,163 1 ,000 2318 2146 2504
No grant ,122 ,017 51,254 1 ,000 1130 1092 1168
Foreign nationality ,196 ,039 25,052 1 ,000 1216 1127 1313
Not working −,387 ,033 140,999 1 ,000 ,679 ,637 ,724
Constant −1954 ,048 1623,868 1 ,000 ,142
Reference category: student at the University of La Laguna, female, 25 years or older, Social and Legal Sciences, with an entrance grade of >10, Spanish nationality and
working.
M.O. Gonz´
alez-Morales et al.
Acta Psychologica 252 (2025) 104669
5

those enrolled in the Science and Arts and Humanities streams. These
authors also found that the students with the lowest probability of
dropping out were those studying Health Sciences degrees. Thus, as
stated, H2.1 was not fullled. However, it was conrmed that the lower
the university entrance qualication, the more likely students are to
drop out. As Rodríguez-Pereiro et al. (2019) point out, prior attainment
is a crucial factor that has a decisive inuence on students' later
achievement. These results are in line with those of Casanova et al.
(2018), who pointed out that low student performance is one of the
factors with the highest incidence in academic dropout. This is a rele-
vant fact, with a clear practical projection, as it highlights the need to
provide special attention to these students, so that they can integrate in
better conditions at university.
In terms of economic characteristics, not having a scholarship/grant or
combining work and studies have an inuence on dropout, given that
students without a scholarship and those who are working are the most
likely to drop out. These results coincide with the ndings in other
research (Fern´
andez-Martín et al., 2019; Constante-Amores et al., 2021;
Fern´
andez-Mellizo, 2022), where it has been shown that students with a
lower socioeconomic status have a higher risk of dropping out of
university.
Ultimately, the results obtained in this research are in line with those
reported by Rodríguez-Pineda and Zamora-Araya (2021) and ˇ
Sabi´
c and
Puzi´
c (2022), who also highlighted that it is not possible to speak of a
single factor when explaining the causes of university academic dropout,
but rather a network of factors that intervene to cause the interruption of
studies. Aina et al. (2022, p.16) presented results that point in the same
direction: university dropout is a consequence of “a combination of in-
dividual, institutional and economic factors”. Their effects on the deci-
sion to discontinue training are mediated by the students' inability to
integrate into the training system. The work of Schneider and Preckel
(2017) also reects the cluster of variables that are related to school
dropout. Students who are well adjusted to university and perform well
tend to be characterised by a high level of self-efcacy, a high degree of
prior achievement and intelligence and high levels of responsibility, as
well as application of appropriate learning strategies to achieve desir-
able goals. In turn, these high-achieving students tend to have faculty
who work hard to design courses well, set clear learning objectives and
provide feedback practices. A recent study by Paseggia et al. (2023)
examined the relationships between motivation to study, retrospective
evaluation of school experiences, subjective well-being, academic
engagement and intention to drop out. They tested different structural
equation models to analyse the relationships between variables and all
of them showed that motivational styles predicted students' engage-
ment, which in turn directly and indirectly inuenced their intention to
drop out.
The research we present constitutes an important contribution to the
challenge of improving the situation of the academic problem in uni-
versities. The analyses carried out made it possible to identify various
characteristics (personal, academic, social, etc.) that have an impact on
dropout, which can facilitate decision-making on measures of various
kinds to reduce the impact of dropout. From the results obtained in the
study, it is concluded that it is important to improve the initial infor-
mation systems for students and to facilitate integration into academic
life in the rst year, as these are key to integration and academic success
and would reduce the risk of dropping out. And in this sense, it is
important to strengthen the tutorial and guidance work of university
lecturers, who must not only educate but also accompany students and
guide them in their learning process (´
Alvarez-P´
erez & L´
opez-Aguilar,
2023).
For the future, it would be interesting to complete this research with
other studies of a more qualitative nature, in order to further study
variables such as gender, branch of knowledge, available economic re-
sources, place of origin, professional activity, etc. and their impact on
academic dropout. This would even make it possible to delve deeper into
how students make the decision to drop out, which can provide sensitive
information to be able to act in the future in a preventive manner in the
face of this problem. Works such as those by Lindner et al. (2023), in
which a longitudinal study over three years was used to conrm the
positive relationship between procrastination, satisfaction with studies
and dropout intention in university students, constitute a good meth-
odological reference to be considered in the future.
Future research could also look more closely at other variables that
could also help to explain dropout more precisely, such as motivation,
satisfaction with studies, academic commitment and similar aspects.
Finally, the need to further rene the instruments for collecting data on
academic dropout should be noted.
CRediT authorship contribution statement
María Olga Gonz´
alez-Morales: Writing – review & editing, Writing
– original draft, Methodology, Formal analysis, Data curation. David
L´
opez-Aguilar: Writing – review & editing, Writing – original draft,
Funding acquisition, Formal analysis, Conceptualization. Pedro
Ricardo ´
Alvarez-P´
erez: Writing – review & editing, Writing – original
draft, Conceptualization. Pedro Antonio Toledo-Delgado: Writing –
review & editing, Writing – original draft, Methodology, Data curation.
Funding
This article derives from the R +D +i Project “Analysis of the
explanatory factors for dropping out of university studies and strategic
actions for their improvement and prevention” (PID2020-114849RB-
I00). Awarded by the Ministry of Science and Innovation (2020) of the
Government of Spain.
Declaration of competing interest
The authors of the manuscript entitled “Dropping out of higher ed-
ucation: Analysis of variables that characterise students who interrupt
their studies” declare that they have no conicts of interest.
Data availability
The data that has been used is condential.
References
Acevedo, F. (2021). Concepts and measurement of dropout in higher education: A critical
perspective from Latin America. Issues in Educational Research, 31(3), 661–678.
https://doi.org/10.3316/informit.190795958120286
Ahn, M. Y., & Davis, H. H. (2020). El sentido de pertenencia de los estudiantes y su
estatus socioecon´
omico en la educaci´
on superior: una aproximaci´
on cuantitativa.
Ense˜
nanza en la educaci´
on superior, 1–14. https://doi.org/10.1080/
13562517.2020.1778664
Aina, C., Baicía, E., Casalone, G., & Pastore, F. (2022). The determinants of university
dropout: A review of the socio-economic literature. Socio-Economic Planning Sciences,
79, Article 101102. https://doi.org/10.1016/j.seps.2021.101102
´
Alvarez, D. (2021). An´
alisis del abandono universitario en Espa˜
na: un estudio
bibliom´
etrico. Publicaciones, 51(2), 241–261. https://doi.org/10.30827/
publicaciones.v51i2.23843
´
Alvarez, P., & L´
opez-Aguilar, D. (2020). Competencias de adaptabilidad y factores de
´
exito acad´
emico del alumnado universitario. Revista Iberoamericana de Educaci´
on
Superior (RIES), 11(32), 46–66. https://doi.org/10.22201/
iisue.20072872e.2020.32.815
´
Alvarez, P., L´
opez-Aguilar, D., & Valladares, R. A. (2021). La inuencia del engagement
en las trayectorias formativas de los estudiantes de bachillerato. Estudios Sobre
Educaci´
on (ESE), 40, 27–50. https://doi.org/10.15581/004.40.27-50
´
Alvarez-P´
erez, P. R., & L´
opez-Aguilar, D. (2023). Tutoría y proyecto formativo del
alumnado universitario: la importancia de comenzar bien. Revista Brasileira de
Orientaç˜
ao Prossional, 23(2), 127–137. https://doi.org/10.26707/1984-7270/
2022v23n0202
B¨
aulke, L., Grunschel, C., & Dresel, M. (2021). Student dropout at university: A phase
orientated view on quitting studies and changing majors. European Journal of
Psychology of Education, 37, 853–876. https://doi.org/10.1007/s10212-021-00557-x
Benítez, R. L., Pazmi˜
no Cevallos, M. R., & Herrera Navas, C. D. (2021). Causes that cause
school dropout in high school students. Tse’De, 4(3). https://tsachila.edu.ec/ojs/in
dex.php/TSEDE/article/view/94.
M.O. Gonz´
alez-Morales et al.
Acta Psychologica 252 (2025) 104669
6

Bernardo, A., Esteban, M., Fern´
andez, E., Cervero, A., Tuero, E., & Solano, P. (2016).
Comparison of personal, social and academic variables related to university dropout
and persistence. Frontiers in Psychology, 7, 1–9. https://doi.org/10.3389/
fpsyg.2016.01610
Bertola, G. (2023). University dropout problems and solutions. Journal of Economics, 138,
221–248. https://doi.org/10.1007/s00712-022-00814-7
Calamet, F. (2020). Factores explicativos del abandono de los estudios en la educaci´
on
superior en contextos socio-acad´
emicos desfavorables. Revista Espa˜
nola de Pedagogía,
78(276), 253–270. https://doi.org/10.22550/REP78-2-2020-02
Casanova, J. R., Cervero, A., Nú˜
nez, J. C., Almeida, L. S., & Bernardo, A. (2018). Factors
that determine the persistence and dropout of university students. Psicothema, 30(4),
408–414. https://doi.org/10.7334/psicothema2018.155
C´
espedes-L´
opez, M. F., Mora-García, R. T., P´
erez-S´
anchez, R., & P´
erez-S´
anchez, J. C.
(2021). El abandono de los estudios universitarios en ense˜
nanzas t´
ecnicas: un caso
de estudio. In R. Satorre (Ed.), Nuevos retos educativos en la ense˜
nanza superior frente
al desafío COVID-19 (pp. 485–498). Octaedro.
Comisi´
on Europea, Direcci´
on General de Educaci´
on, Juventud, Deporte y Cultura,
Wollscheid, S., Stensaker, B., & Jongbloed, B. (2015). Abandono y nalizaci´
on de la
educaci´
on superior en Europa: informe principal. Ocina de Publicaciones. https://doi.
org/10.2766/826962
Constante-Amores, A., Florenciano Martínez, E., Navarro Asencio, E., & Fern´
andez-
Mellizo, M. (2021). Factores asociados al abandono universitario. Educaci´
on XX1, 24
(1), 17–44. https://doi.org/10.5944/educXX1.26889
CRUE. (2022). La universidad espa˜
nola en cifras: 2019/2020. CRUE.
Cuji, B., Gavilanes, W., & P´
erez Constante, M. (2023). Predictive model of student
dropout based on logistic regression. ESPOCH Congresses: The Ecuadorian Journal of
S.T.E.A.M., 3(1), 630–656. https://doi.org/10.18502/espoch.v3i1.1447
De la Cruz-Campos, J. C., Victoria-Maldonado, J. J., Martínez-Domingo, J. A., & Campos-
Soto, M. N. (2023). Causes of academic dropout in higher education in Andalusia
and proposals for its prevention at university: A systematic review. Frontiers in
Education, 8. https://doi.org/10.3389/feduc.2023.1130952
De Silva, L. M. H., Chounta, I. A., Rodríguez-Triana, M. J., Roa, E. R., Gramberg, A., &
Valk, A. (2022). Toward an institutional analytics agenda for addressing student
dropout in higher education: An academic Stakeholders’ perspective. Journal of
Learning Analytics, 9(2), 179–201. https://doi.org/10.18608/jla.2022.7507
Díaz, P., De Le´
on, A. T., & Saavedra, L. M. (2021). Models of analysis and prevention of
university student dropout aimed at the Panamanian context. South Florida Journal of
Development, 2(2), 1349–1357. https://doi.org/10.46932/sfjdv2n2-018
Eicher, V., Staerkl´
e, C., & Cl´
emence, A. (2014). I want to quit education: A longitudinal
study of stress and optimism as predictors of school dropout intention. Journal of
Adolescence, 37(7), 1021–1030. https://doi.org/10.1016/j.adolescence.2014.07.007
Esteban, M., Bernardo, A., Tuero, E., Cervero, A., & Casanova, J. (2017). Variables
inuyentes en progreso acad´
emico y permanencia en la universidad. European
Journal of Education and Psychology, 10(2), 75–81. https://doi.org/10.1016/j.
ejeps.2017.07.003
Fern´
andez-Martín, T., Solís-Salazar, M., Hern´
andez-Jim´
enez, M. T., & Moreira-
Mora, T. E. (2019). Un an´
alisis multinomial y predictivo de los factores asociados a
la deserci´
on universitaria. Revista Electr´
onica Educare, 23(1), 73–97. https://doi.org/
10.15359/ree.23-1.5
Fern´
andez-Mellizo, M. (2022). An´
alisis del abandono de los estudiantes de grado en las
universidades presenciales en Espa˜
na. Ministerio de Universidades.
Freitas, P., Martins, I., Noronha, D., Noronha, A., Cruz, C., V´
azquez, E., & Costa, C.
(2022). Dropout factors in higher education: A literature review. Psicología Escolar e
Educacional, 26, 1–10. https://doi.org/10.1590/2175-35392022228736T
Freixa, M., Llanes, J., & Venceslao, M. (2018). El abandono en el recorrido formativo del
estudiante de ADE de la Universidad de Barcelona. Revista de Investigaci´
on Educativa,
36(1), 185–202. https://doi.org/10.6018/rie.36.1.278971
Gairín, J., Triado, X. M., Feixas, M., Figuera, P., Aparicio-Chueca, P., & Torrado, M.
(2014). Tasas de abandono estudiantil en las universidades catalanas: perl y
motivos de desconexi´
on. Calidad en la Educaci´
on Superior, 20(2), 165–182. https://
doi.org/10.1080/13538322.2014.925230
Gallego, M. G., P´
erez, A. P., & Gallego, J. C. G. (2021). Identicaci´
on de estudiantes en
riesgo de abandono acad´
emico en la educaci´
on superior. Ciencias de la Educaci´
on, 11,
427. https://doi.org/10.3390/educsci11080427
Hern´
andez, A., & Vargas, E. (2016). Condiciones del trabajo estudiantil urbano y
abandono escolar en el nivel medio superior en M´
exico. Estudios demogr´
acos y
urbanos, 31(3), 663–696. http://www.scielo.org.mx/scielo.php?script=sci_arttext
&pid=S0186-72102016000300663&lng=es&tlng=es.
Huo, H., Cui, J., Hein, S., Padgett, Z., Ossolinski, M., Raim, R., & Zhang, J. (2023).
Predicci´
on de la deserci´
on escolar para estudiantes universitarios no tradicionales:
un enfoque de aprendizaje autom´
atico. Revista de retenci´
on de estudiantes
universitarios: investigaci´
on, teoría y pr´
actica, 24(4), 1054–1077. https://doi.org/
10.1177/1521025120963821
Kehm, B. M., Larsen, M. R., & Sommersel, H. B. (2019). Student dropout from universities
in Europe: A review of empirical literature. Hungarian Educational Research Journal, 9
(2), 147–164. https://doi.org/10.1556/063.9.2019.1.18
Li, I. W., & Carroll, D. R. (2020). Factors inuencing dropout and academic performance:
An Australian higher education equity perspective. Higher Education Policy and
Management, 42(1), 14–30. https://doi.org/10.1080/1360080X.2019.1649993
Lindner, C., Zitzmann, S., Klusmann, U., & Zimmermann, F. (2023). From procrastination
to frustration—How delaying tasks can affect study satisfaction and dropout
intentions over the course of university studies. Learning and Individual Differences,
108, Article 102373. https://doi.org/10.1016/j.lindif.2023.102373
Lizarte, E., & Gij´
on, J. (2022). Prediction of early dropout in higher education using the
SCPQ. Cogent Psychilogy, 9, 1–13. https://doi.org/10.1080/
23311908.2022.2123588
Marco-Franco, J. E. (2022). Impact of marine sound pollution from merchant ships (Doctoral
thesis).
Matta, C. M. B., Lebr˜
ao, S. M. G., & Heleno, M. G. V. (2017). Adaptaç˜
ao, rendimento,
evas˜
ao e vivˆ
encias acadˆ
emicas no ensino superior: Revis˜
ao da literatura. Psicologia
Escolar e Educacional, 21(3), 583–591. https://doi.org/10.1590/2175-
353920170213111118
Medel-Ramírez, C., & Medel-L´
opez, H. (2018). Complementarity analysis of the priority
areas development program and the priority attention areas program in the National
Crusade against Hunger Program in indigenous municipalities in the state of
Veracruz Mexico. ECORFAN-Mexico Journal, 9(20), 29–44. https://philpapers.org
/archive/MEDCAO.pdf.
Merino-Tejedor, E., Hontangas, P. M., & Boada-Grau, J. (2016). Career adaptability and
its relation to self-regulation, career construction, and academic engagement among
Spanish university students. Journal of Vocational Behavior, 93, 92–102. https://doi.
org/10.1016/j.jvb.2016.06.001
Ministry of Universities. (2022). Datos y cifras del Sistema Universitario Espa˜
nol.
Publicaci´
on 2021-2022. Disponible en. https://www.universidades.gob.es/wp-conte
nt/uploads/2022/11/Datos_y_Cifras_2021_22.pdf.
OECD. (2016). Education in Colombia. Reviews of National Policies for Education. OECD
Publishing.
Parra-S´
anchez, J. S., Torres, I. D., & Martínez, C. Y. (2023). Factores explicativos de la
deserci´
on universitaria abordados mediante inteligencia articial. Revista Electr´
onica
de Investigaci´
on Educativa, 25, 1–7. https://doi.org/10.24320/redie.2023.25.
e18.4455
Paseggia, R., Testa, I., Esposito, G., De Luca Picione, R., Ragozini, G., & Freda, M. F.
(2023). Examining the relation between rst-year university Students’ intention to
drop-out and academic engagement: The role of motivation, subjective well-being
and retrospective judgements of school experience. Innovative Higher Education, 48,
837–859. https://doi.org/10.1007/s10755-023-09674-5
Pe˜
na-V´
azquez, R., Gonz´
alez Morales, O., ´
Alvarez-P´
erez, P. R., & L´
opez-Aguilar, D.
(2023). Building the prole of students with the intention of dropping out of
university studies. Revista Espa˜
nola de Pedagogía, 81(285), 291–315. https://doi.org/
10.22550/REP81-2-2023-03
P´
erez, F., & Ald´
as, J. (2019). Indicadores sint´
eticos de las universidades espa˜
nolas. Valencia:
Fundaci´
on BBVA-Instituto Valenciano de Investigaciones Econ´
omicas. https://www.
fbbva.es/wp-content/uploads/2019/04/Informe-U-Ranking-FBBVA-Ivie-2019.pdf.
P´
erez, P., P´
erez, H., & Guevara, G. (2022). Factores de riesgo y desarrollo de resiliencia
en adolescentes. Revista Cientíca UISRAEL, 9(2), 23–38. https://doi.org/10.35290/
rcui.v9n2.2022.519
Piepenburg, J. G., & Beckmann, J. (2021). The relevance of social and academic
integration for students’ dropout decisions. Evidence from a factorial survey in
Germany. European Journal of Higher Education.. https://doi.org/10.1080/
21568235.2021.1930089
Pusztai, G., F´
enyes, H., & Kov´
acs, K. (2022). Factors inuencing the chance of dropout or
being at risk of dropout in higher education. Education Sciences, 12(11), 804. https://
doi.org/10.3390/educsci12110804
Rodríguez-Pereiro, S., Regueiro, B., Rodríguez, S., Pi˜
neiro, I., Est´
evez, I., & Valle, A.
(2019). Rendimiento Previo e Implicaci´
on en los Deberes Escolares de los
Estudiantes de los Últimos Cursos de Educaci´
on Primaria. Psicología Educativa, 25,
109–116. https://doi.org/10.5093/psed2019a2
Rodríguez-Pineda, M., & Zamora-Araya, J. (2021). Abandono temprano en estudiantes
universitarios: un estudio de cohorte sobre sus posibles causas. Uniciencia, 35(1),
19–37 (https://dx.doi.org/10.15359/ru.35-1.2).
ˇ
Sabi´
c, J., & Puzi´
c, S. (2022). Exploring dropout risk in higher education in Croatia: An
empirical analysis. Issues in Educational Research, 32(3), 1153–1173. http://www.
iier.org.au/iier32/sabic.pdf.
Schneider, M., & Preckel, F. (2017). Variables associated with achievement in higher
education: A systematic review of meta-analyses. Psychological Bulletin, 143(6),
565–600. https://doi.org/10.1037/bul000009
Thuy, T., Kaur, A., & Busthami, A. H. (2017). A self-determination theory based
motivational model on intentions to drop out of vocational schools in Vietnam.
Malaysian Journal of Learning and Instruction, 14(1), 1–21. https://doi.org/10.32890/
mjli2017.14.1.1
Tuero, E., Cervero, A., Esteban, M., & Bernardo, A. (2018). Why do university students
drop out? Inuencing variables regarding the approach and consolidation of drop
out. Educaci´
on XX1, 21(2), 131–154. https://doi.org/10.5944/educxx1.20066
Zamora, M. A., Moreira, E. W., Espinoza, K. P., & Daza, M. Y. (2023). Analysis of the
socioemotional factors incident in the student dropout rate of nivelaci´
on-ESPAM.
Maestro y Sociedad, 103–112. https://maestroysociedad.uo.edu.cu/index.php/MyS/
article/view/6029.
M.O. Gonz´
alez-Morales et al.
Acta Psychologica 252 (2025) 104669
7