Student dropout rates in Catalan universities: profile and motives for disengagement

Data from over 21,600 students who left Catalan higher education institutions during the academic years 2000–2001 and 2001–2002 have been analysed in order to describe the academic and personal profiles of university dropouts. Additionally, a telephone survey and face-to-face interviews with a pilot group of leavers were conducted to gather additional qualitative information about the reasons for their decision. The influences on non-completion can be reduced to three main factors, among which dissatisfaction with the quality of the students’ experience, family and work responsibilities as well as economic difficulties, are prominent. Analysis conducted showed no significant differences between Catalan higher education institutions, with dropout percentages ranging from 28% to 33%. These rates should be understood within a context of broad access to higher education. Within the European Higher Education Area scenario, the findings have triggered institutional endeavours to improve the quality of the students’ and teachers’ experience.


Introduction
The Bricall (2000), Attali (1998), National Committee of Inquiry into Higher Education (NCIHE) (1997) and Mandelson (2009) reports on the Spanish, French and British educational systems, respectively, highlighted the dropout issue in the 1990s as it had a bearing on the effectiveness and efficiency of the university system. They reaffirmed the idea that the phenomenon of university dropout rates had to be reduced in order to increase the quality standards and productivity of the institution as well as student satisfaction. Since the 1990s, the dropout rate has been quantified and used as a performance indicator. Student dropout has been a widely discussed topic among academic authorities in the European Union (EU), within each country and institution, and has increased with the adaptation to the European Higher Education Area. However, actions encouraging student retention have still not been assessed.
Dropout takes on broader dimensions when extended, almost universally, to access to university studies. Behind this problem lies the economic cost of university studies available to the majority of students and the inefficient use of resources, as well as the problems of dissatisfaction for those who wanted to be educated but whose expectations were not fulfilled or were not properly and professionally oriented towards the type of goals they could hope to fulfil. Undoubtedly, the challenge facing the systems is to maximise the development of a country's human resources through its higher education institutions.
This article addresses the analysis of student dropout rates in public universities in Catalonia, Spain. The Catalan university system is composed of 12 universities (five of which are private) with a total of 225,181 undergraduate students enrolled (3.9% of the population) (Catalan University Quality Assurance Agency (Agència per a la Qualitat del Sistema Universitari de Catalunya, AQU), 2007). It is also worth mentioning the considerable drop by 7.7% of new-entry students over five consecutive years (2000)(2001)(2002)(2003)(2004)(2005) caused by demographic factors. According to data for the entire student population of Catalan public universities, during the 2004-2005 academic year, 76% of new university students were enrolled in their first-choice degree, with social sciences being the area with highest enrolments (47%) and demand (48%).
More specifically, the aim is to investigate the leavers' profile and the factors that may explain or shed new light on what motivates dropout in the Catalan university system. The ultimate goal is to identify strategies that will help to guarantee greater student retention rates and success. Data from over 21,600 students who left Catalan higher education institutions during the academic years 2000-2001 and 2001-2002 are analysed and contrasted with telephone and face-to-face interviews with dropout students. Beyond the interpretations that can be applied to the dropout rate and the models for analysing it, there are many questions to be addressed. What is really known about what takes place at university? What motivates some students to carry on studying? What causes students to drop out? Which of these causes or factors are universities responsible for? How can dropout rates be measured? Is there a type of student profile that is more likely to dropout? Can an explanatory, integrated model for this phenomenon be found? These are some of the questions that this paper attempts to address based on the outcomes of interdisciplinary research financed by the AQU.
Towards a conceptualisation of the phenomenon A precise, conceptual definition of university dropout is a complex task that goes beyond the theoretical arena and manifests itself in the policies, actions and studies undertaken by universities and countries all over the world. It is also hard to measure because it requires not only knowing what it is that is to be measured but also having suitable and precise institutional data, collected systematically over a certain period of time.
Any bibliographical review displays a huge variation in terminology, which hampers comparative studies and definitions of the descriptors that must be considered for further analysis. For this reason, the concept of dropout and related terms (desertion, stopping, completion, discontinuity, (non-) persistence, survival or retention) depend substantially on the context and research objectives devised (Berger & Lyon, 2005;Mortenson, 2003). It also depends on whether dropping out is to be considered a failure or not, as well as on the transition model in which the process is framed (Tinto, 2010). Longden (2001) synthesised and classified all the terminological casuistry. Of all the terms in use, the most common in the literature in English are ʻdropoutʼ, ʻattritionʼ and ʻwithdrawalʼ, which in the Catalan context are referred to as abandonament. The use of the terms academic failure and dropout or desertion as synonymous is being replaced by a terminology that gives room to the several types of dropout. Thus, Bourdages (1996) introduced the idea of focusing on persistence rather than investigating dropout, and non-persistence rather than retention or study continuation.
The National Association of Universities and Higher Education Institutions (Asociación Nacional de Universidades e Instituciones de Educación Superior) (ANUIES, 1986) described dropping out as not attending class and the non-fulfilment of obligations in such a way as to affect the ultimate effectiveness of university degree performance. The dropout rate can also be defined as the percentage number of students in a cohort that have not completed their degree in an established period of time.
Additional research demonstrates that there are problems with students who, while they do not drop out, repeat courses, change institution to enrol in the same degree or do not attend courses regularly or take examinations. Thus, it is a challenge to find an accurate definition including all possible typologies.
For instance: For Tinto (1975), dropouts are the flow of students who definitively abandon all forms of university education, or the moment at which students undertake immediate transfers to another institution. García Areito (1986) gives three types of definitions: ʻnon-startersʼ, or those who have never been assessed despite being enrolled for one or two years; real dropouts, referring to students who are examined before leaving the degree; and global dropouts, as the combination of the two aforementioned concepts. Altamira (1997) describes desertion from four points of view: voluntary and definite dropout; dropout by expulsion due to poor academic performance; dropout by changing study programmes; and dropout from disciplinary expulsion.
It is necessary, therefore, to bear in mind the different types of dropout: voluntary, involuntary, temporary, permanent, initial, provisional, definite and the possible relationship (or not) between dropping out and academic failure, or dissatisfaction with the quality of the student's experience (Yorke & Longden, 2008).
It is also necessary to take into account more precise casuistry, such as enrolling for a degree and not sitting examinations, which Himmel (2002) terms 'premature baby' whereas Giovagnoli (2002) terms it 'epidemic', when referring to those 'absent' students who do not obtain the necessary credits to continue studying simply because they have stopped going to university.
In order to be able to analyse the dropout phenomenon and to facilitate its calculation, in the context of this research, dropout in Spain means students who registered during a course and did not formally enrol again for the next two course periods. Operative definitions used since the 1990s in the evaluation processes of university quality by the Ministry of Education (Catalogue of Indicators in the Spanish Public University System) (Consejo de Coordinación Universitaria, 2002) also associate university dropout with performance indicators. They have also been used as indicators in higher education institutional rankings. These indicators have evolved in parallel with the technological development of universities for gathering and systematising a whole range of information in their databases.
Explanatory models of university dropout University dropout appeared as an object of study in the 1960s and by the 1970s a theoretical body referred to Tinto's interactionist theory (Berger & Lyon, 2005). Since then, most literature on university study continuity has developed along two lines: the Student Integration Model by Spady (1970) and Tinto (1975); and the Student Attrition Model by Bean (1980). Tinto's model has evolved over time and has encouraged the development of an important part of later theoretical models that attempt to comprehend a clearly complex and multidimensional phenomenon.
Among the classification models, those of the following are most notable: (a) the classification by Braxton et al. (1997), revised by Himmel (2002) and, recently, by Donoso and Schiefelbein (2007); (b) that of Cabrera et al. (2006); and (c) that of Southerland (2006). Table 1 illustrates the coincident and divergent elements of the three classifications of the explanatory models for the university dropout phenomenon.
Taking the Braxton et al. (1997) classification as a reference, the nonpersistence analysis approaches can be grouped into five broad categories (psychological, sociological, economic, organisational and integrative) according to the phenomenon's explanatory variables, whether individual, institutional or stemming from the family environment.
More specifically, Tinto's model is based on the fact that academic and social integration explains permanence in the educational system and is influenced, on the one hand, by the student's cultural knowledge on entry, arising from their previous academic background, family environment and personal characteristics, and, on the other hand, by the initial commitment to the institution and the intention of completing the degree, as well as positive interactions with the environment (participation in extracurricular activities). Tinto is the first author to insist on several dropout typologies: the voluntary and the involuntary or normative dropout. Pascarella and Terenzini (2005), extending Tinto's model, considered the factors that directly or indirectly influence students' academic performance and the institution's characteristics. Subsequent studies have demonstrated that these factors have a powerful influence on those sectors that are in the minority. Bean (1983Bean ( , 1995 added a psychological perspective: a person's attitudes and behaviour and external factors have a major influence on persistence (family support, family responsibilities and economic resources). Bean incorporates the characteristics of the productivity model developed in the context of work organisations into Tinto's model. According to Bean, the following factors have an effect on desertion: (a) academic factors: prior to university, academic integration and academic results; (b) psycho-social factors: goals, perceived utility, interaction with parents and teachers; (c) environmental factors: funding, opportunities to change university, external social relations; (d) socialisation factors: academic performance, adaptation and institutional commitment.
Additionally, Bean and Vesper (1990) observed that non-cognitive factors such as personal (attitudes, aspirations, motivations, interests), environmental and organisational characteristics also significantly influence desertion and particularly, voluntary desertion.
In an extension to his original model, Tinto integrated the previous theoretical contributions and transferred them to his current persistence model.
In order to incorporate Bean's contributions, Cabrera et al. (1993) and St. John et al. (2000) set out a holistic model that entailed three phases for the student. In a first phase, academic skills, previous experiences and socio-economic factors influence the decision to continue one's studies. Second, students value the degree's cost-benefit aspect and therefore make an initial commitment to complete their studies. In the final stage, after entering a degree course, other factors come into play that modify or reinforce initial aspirations.

Methodological approach
Dropout complexity in higher education, from the point of view of theoretical models, and considering the lack of coincidence in assessing the phenomenon in the international context (Feixas et al. 2010), underlines the need to conduct a study in the Catalan university context. This analysis is performed in two research phases, using a macro and micro approach.
The first phase analyses the dropout phenomenon in Catalonia. Using the student data collected from the UNEIX platform (the Portal of the Information System for Universities and Research in Catalonia) a descriptive analysis of the population has been made, facilitating first a quantitative and second, a qualitative approach to the dropout rate.
The database provided by the Catalan University Quality Assurance Agency, with the permission of the Directorate-General of Universities of the Generalitat de Catalunya, gathers information on the 2005-2006 academic year dropouts made up of those students who entered university in the academic years 2000-2001 and 2001-2002, amounting to a total of 21,620 dropouts for the whole Catalan university system.
The following variables are contained in the database provided (Table 2). In the second phase, a micro approach from a pilot sample was applied to the specific situation of dropout students in our universities through a descriptive and comprehensive study via telephone and face-to-face interviews. This helped to distinguish the reasons why these students gave up their studies and also gathered feedback to set identifiable retention strategies. The bibliographical review performed highlights the absence of any standardised system facilitating the study or awareness of the causes leading students at any particular moment to decide not to continue their degree. For this reason, a survey 'Cuestionario para evaluar el perfil del alumnado universitario que abandona sus estudios' [Questionnaire to assess the dropout university students' profile] ) and interview protocols were designed. Both include all the theoretical variables that the aforementioned models consider relevant to explain university study dropout (Table 3). In the data collection phase, this exploratory research, of a descriptive and integral nature, included additional quantitative and qualitative elements from the methodological standpoint.
The survey, previously validated by experts, consisted of 65 items. Except for the personal and academic data, in most questions the student expressed an opinion on a scale corresponding to a Likert 1-to-10 scale.
Decisions made regarding the sample were related to the data facilitated by AQU Catalonia. The theoretical sample comprised 801 students, the total number of students commencing university studies in 2000-2001. A dropout analysis was performed by degrees and areas and the sample was therefore selected through a proportional stratified sampling taking into account student dropout rates in Catalan public universities and the data on new entry students, to calculate the dropout rates per degree. Five degrees (one for each area of knowledge) were selected to guarantee a balanced representation across all the sampling variables (including degrees, number of students, universities): business administration and management, biology, computer engineering, history and medicine.
After selecting the degrees, localisation and selection of the pilot sample were addressed. The process was as follows: Identification and selection of universities: Universitat Autònoma de Barcelona, Universitat de Barcelona, Universitat de Girona, Universitat Politècnica de Catalunya, Universitat Pompeu Fabra and Universitat de Lleida. Institutional request for list of dropouts in the selected degrees. Once available, a random sample was made, taking into account a proportionate distribution by students and degrees.
The effective student sample for telephone survey was 275. The response rate was 35.63% with a global error margin of +/-5.3% and a confidence level of 95.5%. The study was carried out during the period September to December 2008.
Before performing the telephone surveys, a group of pollsters was hired and trained. The Internet was also used for data collection because some telephone surveys proved impossible.
Additionally, in order to further explore dropout reasons and complement the survey's information, a series of personal semi-structured interviews averaging an hour in length was performed. A sub-sample was taken following the same criteria as with the telephone surveys. All students interviewed were previously polled to gauge their willingness to collaborate. The theoretical sample size was 30 students but the actual number was 17, interviewed in November and December 2008.

Analysis and discussion of results
The results of the macro analysis after database examination are shown as well as the micro analysis focusing on the telephone survey and personal interviews.
It is worth mentioning the fact that all Catalan universities face a dropout problem, which averages out at 33.6%. In ranking order, 37% of student dropouts are from the Universitat Politècnica de Catalunya, followed by the Universitat de Barcelona, the Universitat Autònoma de Barcelona and, finally, the Universitat de Lleida at 28%, although in relative terms, the higher the enrolment figures, the higher the dropout rate.
Degree choice further reveals the differences. Dropout rates by degree choice oscillate between 20% and 60%, and are even higher in certain engineering courses or lower in less traditional subjects such as physiotherapy.
Dropout rates by degrees also vary due to differences in discipline typology, teaching and learning culture, student ratios, facilities, student support, satisfaction, motivation and academic quality. While recognising the wide range of variation between degrees, highest dropout rates are found in social sciences (40%), technology (29.6%) and arts and humanities (17.2%). About half of those who drop out do so in their first year of study at Catalan universities, considering that dropout out means leaving the degree but not the university. This corroborates what other research has shown: the importance to look after the students during their first-year experience (Yorke & Longden, 2008). The average period spent by a dropout at university is a little over one and a half years.
The student's profile is informed by different variables, which can be grouped into three different moments regarding the student's degree: (1) academic and personal variables at the beginning of the degree; (2) the type of degree the student has abandoned; and (3) university re-entry policy.
(1) Student's situation at beginning of degree: socio-demographic and social variables and previous academic variables Regarding personal features, a dropout's socio-demographic profile shows little clear differentiation as to age, gender, parents' studies or employment. Dropout affects different age groups, more men than women (51.4% compared to 48.6%, respectively) and cannot be related to the parents' study level or employment. In this regard, the student's socio-demographic profile is not a determining factor in the dropout phenomena.
Most of the students drop out during their first year and mainly in technology studies. This invites one to consider the processes and mechanisms of academic promotion that different universities and degrees regulate.
As regards the academic aspects, the highest numbers of dropouts were students who accessed university from secondary school and took the Selectivity test (Spanish examination at the end of secondary school studies that allows access to university), while there are fewer dropouts from basic vocational training and in higher vocational degrees. Among the former were students from the arts and social sciences Baccalaureate. Another significant percentage is that of students sitting the entrance examination for those aged 25 and above (14.17%).
(2) Variables affecting the abandoned degree, such as academic performance Students who drop out usually have a poor academic record: credit points achieved (European Credit Transfer System (ECTS)) are about 50% of those required to be considered as a student with optimum academic performance. More specifically, dropouts in technical and experimental science degrees gain fewer ECTS, unlike in health science degrees, where students have a better academic record.
(3) Ex-post situation, considering the change of degree and university re-entry As regards re-entering other degree studies in Catalonia, the analysis shows that over 60% of students who abandoned their studies in [2000][2001] have not re-entered the university system. It is likely that a percentage of these students continued their studies outside Catalonia, because a significant percentage of them took the Selectivity tests in other Spanish regions. In addition, most students re-entering the university system are those who dropped out of technology and health science degrees.
The key question in the questionnaire about the main reason behind university dropout assesses the two types of factors to which students may attribute their decision to drop out. In keeping with the theory, the ʻinternalʼ factors have been defined as a lack of motivation and unfulfilled expectations, and the ʻexternalʼ factors are those that are outside the control of the students (timetables, family responsibilities and so on). The results reveal that 44.8% of the sample attribute the decision to drop out to internal factors (Table 4) and that the type of dropout is directly linked to the factors Integration at university and opportunities for friendship 6.86 6.00 2.37 (sig.019) mentioned by the students; thus the lack of motivation and unfulfilled expectations are actual characteristics of the students who voluntarily decide to leave their studies (Table 5). Variables such as age, typology of students (full-time students and working students) and re-entry to the university are statistically significant variables to define the profile of each sub-sample according to the factors attributed to dropout. The average age of those students who dropped out because of ʻinternalʼ factors is 26.63 (sig. 0.000), they are full-time students (52% of cases, sig 0.001) and have re-entered university (52%, sig 0.021). The profile of those students who dropped out because of 'external' factors is older (average age of 29.29), they combine studies and work in 69% of cases and three out of four have re-entered university.
Research limitations lie in the insufficient indicators or variables in the database which make it difficult to establish relationships among dropout causes as well as limit the possibilities of comparing these data with data from global university enrolment. For these reasons, research conclusions can only underline the overall dropout student profile and outline the percentages and typology of the degrees they abandoned. It is also of interest to know about re-entry students, although we can only tabulate those re-entering the Catalan university system.
For this reason it was deemed necessary to develop new tools to measure the dropout rate in all its complexity. In order to do this, additional data has been collected which is linked to: (a) other academic variables such as learning processes and learning pace, the academic and vocational guidance available, degree motivation, degree evaluation; (b) variables affecting the relationship between the student and the institution and the rationale behind his/her dropping out; (c) institutional variables such as academic support and degree guidance received at university; and (d) personal variables such as employment, the student's economic situation and socialisation factors.
The drafting of two original instruments (a survey and an interview protocol) facilitated this complementary research. On the one hand, the survey examined the profile of those who abandoned their university studies and, on the other hand, the interviews increased awareness of the circumstances leading to the dropout and its causes.
The results inform us about entry into a university degree programme and the fact that students chose their degree mainly out of personal interest, based on guidance and support from relatives and peers. The university choice was usually justified by geographical proximity factors, while aspects such as the prestige and educational policies of each institution were secondary.
Among the reasons for dropping out that emerged from the interviews, university leavers pointed out a lack of motivation and unfulfilled expectations. This feature has a direct effect on academic integration, fundamentally involving a reduction in class attendance and social activities, and is named as a factor contributing to academic success or failure. Moreover, the fact that the decision to leave was mainly recorded during the first semester reinforces the importance of retention strategies focusing on the first-year experience.
Finally, more than half of those who gave up their studies had difficulties reconciling academic activities and employment. The vast majority worked in jobs with little or no connection with the studies. In most cases, dropouts were voluntary.

Concluding remarks
In the international context there is a wide body of knowledge, developed under the protection of universities' interest in optimising retention levels and the excellence of their studies, which has highlighted some key principles.
First, the conceptual and operational complexity of the processes of persistence, abandonment and graduation and the existence of situations or typologies resulting from different situations. A categorisation of these concepts has been put in place in accordance with temporal variables (early dropout, first-year dropout, late desertion and so on), continuity or change of institution variables and duration variables. Second, there is an abundance of theoretical approaches, some of which have lasted over time and which have reached a higher level of scientific and experimental formulation, such as Tinto's explanatory theoretical model (Tinto, 2010).
The OECD's 2009 report maintains that not finishing a university degree does not necessarily mean failure, if the student benefits from the time devoted to the programme. The OECD annual reports, alongside other aforementioned studies, also warn of high dropout rates and the unavoidable political, economic, social and personal consequences.
Today there is ample access to university studies, yet an important increase in the number of students who abandon their studies is also manifest. It is important to analyse the dropout phenomenon from a geopolitical and sociological standpoint, since different countries define and measure it differently.
In Europe, dropout rates in higher education have increased: more than 30% of those enrolled do not complete their degree. The survival rate (calculated from the number of graduates and the number of newly enrolled students) is 71% and Spain tops this by 3% (OECD, 2008).
This research attempts to provide an answer, for Catalonia, the rest of Spain and other countries, to these issues, without underestimating the doubts that still emerge from the study's conclusions.
The macro and micro dropout analysis and therefore that of academic persistence in the Catalan university system, has considered as significant variables for influencing dropout: those evident before leaving the course (socio-demographic and academic), as well as those degrees with the highest dropout rates and the student's situation after leaving university.
The generic analysis demonstrates that since dropout, as a social phenomenon, affects all types of degrees, there are no significant differences between Catalan universities and dropout is slightly less than in Spanish universities (Hernández, 2008). However, differences between study programmes are encountered: the dropout distribution corresponds mainly to social science studies, somewhat less to technology studies and much less to humanities and health science studies and between Catalan and Spanish universities.
The specific analysis based on the students' profile demonstrates that students are less persistent in their studies due to employment reasons and economic situations. Men are less persistent than women. Social background does not appear to have a great influence on dropout rates; however, the high percentage of missing answers in the survey means no definitive conclusions can be drawn. On the other hand, dropout affects to a greater extent those students who had a lower access grade. Finally, it may be stated that the dropout rates by degrees fluctuate between 20% and 60% but their analysis is very complex due to the enormous variety of degrees taken into account.
From the analysis of the academic dynamics, in quantitative terms it can be ascertained that there is more dropout among students coming from other autonomous communities; that dropout and low academic performance are correlated; and finally, that more than 60% of the students who give up their studies do not re-enter the university system.
The macro-and micro-scale analyses are indicative but, at the same time, are limited by the fact that the available data does not facilitate the creation of more and better indicators, nor does it enable it to be compared with the students' global enrolment.
The results of the field study alongside the survey and the interview have highlighted the importance of encouraging these types of studies to provide arguments explaining why students do not continue with their studies and the need for a broader sample and holistic set of variables. Dropout motives are diverse and especially rooted in: (a) the lack of motivation in the degree and, therefore, the subsequent difficulty in fully integrating into academic and university life; (b) difficulties in reconciling academic life with outside employment; (c) lack of economic independence; (d) the fact that institutions do not offer interesting programmes from a professional and methodological point of view; (e) lack of services to support the students' learning and integration processes; and (f) lack of relevance to the reality of the business and employment world.
The specific histories analysed help to gain an understanding of the circumstances accompanying the students' decision to drop out, whether voluntary or imposed. In this regard, the decision to give up their studies focuses on a concrete period of the students' academic life (58% of students who abandon their studies do so in the first year). However, the decision to drop out is rooted in a more complex phenomenon, that is, a set of personal and contextual circumstances that influence the decisions in a significant way.
This study confirms the theory that moments of special relevance exist when facing the decision to drop out and the need to reinforce psychopedagogical attention and personal and academic guidance at university, especially during the first year. Problems deriving from a lack of study motivation and performance might also be related with an inadequate degree selection process which forces one to analyse more critically the guidance processes taking place in post-secondary education.
The flexible organisation of the studies and the implantation of modular systems can have a positive impact on student retention rates. Poor motivation behind the degree is also related to the quality of the academic programme. As a result, it is particularly necessary to review educational methodologies and to ensure individual and social activities with a higher impact.
There is one whole segment of the population that should benefit from higher public loans or scholarships. Financial constraints need to be put to one side as this can have a negative impact on student performance.
Possible re-entry into the university system is also to be considered because interviewees expressed their willingness to re-enter university again to fulfil their university aspirations. This can be seen as an indicator of satisfaction with the institution's quality.
Factors associated with retention are raw associations obtained through cross-tabulation and, therefore, additional insights might be gained through the use of multivariate analysis.
Finally, it is also worth mentioning that the Spanish agenda for the implementation of the Bologna reforms has made great efforts to overcome these difficulties and new study plans and university study structures are incorporating retention measures to overcome the aforementioned situations. These include: the option to study courses part-time, effective interventions to engage, retain and enhance the success of particular students or discipline groups; pre-entry support to engage students and improve their preparation for and transition into higher education; improvements in the design and delivery of the curriculum to promote the success of all students; initiatives and approaches in the academic sphere that promote social engagement with peers and staff; approaches to identify, monitor and engage students at risk of withdrawal; and co-ordination of the student learning experience to promote student engagement, retention and success across programmes, departments, faculties or institutions.