Why Do We Drop Out? Typology of Dropping Out of High School

Theoretical Framework

We used Bronfenbrenner’s (1979) ecological systems theory that describes how different levels of the ecological system (adolescent, family, school, and neighborhood) affect youth’s development. In our work, we used it to examine the factors within systems surrounding high school dropouts. We also used the integrative stress process/life course theory of dropout (Dupéré et al., 2015), a theory that emphasizes the effects of both long-term and short-term stressful life events on school outcomes in young people and the interaction between predisposing long-term risk factors, precipitating stressors, and turning points in eventual high school dropout. The two theories combined guided our exploration of how young people interpret their reasons for dropout across the different systems they live in and enabled us to consider both life course and the proximal experiences that contribute to dropout.

Dropout Typologies

The large body of literature on high school dropout has usually considered individual, family, school, and community factors (e.g., De Witte et al., 2013; McDermott et al., 2019; Rumberger, 2004) in the framework of Bronfenbrenner’s (1979) ecological systems theory. A number of studies have focused on cumulative risk: the more risk factors experienced, the more likely a person will drop out (Freeman & Simonsen, 2015; Hammond et al., 2007). From this perspective, several studies found patterns of factors leading to a typology of high school dropouts (Balfanz et al., 2007; Bowers & Sprott, 2012; Fortin et al., 2006; Janosz et al., 2000; Kronick & Hagris, 1998; Lessard et al., 2008; Menzer & Hampel, 2009). Using a typology can contribute to a deeper understanding of differences between young people who drop out and, therefore, to the design of differential interventions aimed at preventing dropout (Bowers & Sprott, 2012).

Most typology studies have identified four groups, summarized by Bowers and Sprott (2012), as (a) struggling with academics, with the lowest academic achievement and the highest rate of absenteeism and suspensions; (b) bored with the process, with few problems in terms of academic ability but disaffected by school and showing little interest; (c) disruptive of school—the “traditional” dropout (Fortin et al., 2006)—with poor academic achievement and social adjustment difficulties that mean they are the most visible type of dropout; and (d) quiet dropouts, with poor school achievement and little attachment to the school, often with depressive symptoms and often dropping out in response to stressful life events. Bowers and Sprott (2012) described this last subgroup in very broad terms, which possibly explains why it was the largest subgroup of dropouts in their review. Nonetheless, this subgroup probably reflects important nuances that might contribute to a better understanding of young people at risk and the design of more precise interventions. Fortin et al. (2006) has called for more research into students with depressive symptoms because the link between depression and dropout risk has not been sufficiently explored, whereas researchers have begun to pay greater attention to the role played by stressful life events in enhanced dropout risk (Dupéré et al., 2015, 2018; McDermott et al., 2018).

The abovementioned studies on dropout typologies are North American, however, so there is a need for similar studies for other populations, including in Europe. They are also mostly quantitative and there are only few typology studies based on qualitative approach (e.g., Lessard et al., 2008). Qualitative studies could provide more holistic and deeper insights into students’ own perspectives and interpretations of their experience of dropping out (Bunting & Moshuus, 2017; Hunt, 2009).

Our study aims to contribute to bridging those literature gaps by exploring the reasons and experiences of a sample of Croatian high school dropouts and, based on their interpretations of reasons and experiences, proposing a typology of high school dropouts.

Croatia

Croatia joined the European Union (EU) in 2013 and, as a member, is bound by agreements, including the Europe 2020 strategy, which covers educational outcomes, setting, for instance, targets for early school leavers (below 10%) and people aged 30 to 34 years with higher education (minimum 40%; European Commission, 2019). Although Croatia is behind the other EU countries in terms of wealth, education, and social and health benefits, it has a surprisingly low high school dropout rate—just 3.3% in 2018 (Eurostat, 2019a)—although Eurostat points out that recent data for Croatia are unreliable. For comparison, in 2018, countries similar to Croatia in the dropout rate were Slovenia (4.2%) and Switzerland (4.4%), whereas Iceland (21.5%) and Turkey (31%) were on the other side of the bar (Eurostat, 2019a). Eurostat also reports that Croatia was among the few member states where there were more student dropouts who were wanting to work than those who were employed; in the majority of member states, the situation was the opposite (Eurostat, 2019b).

Leaving aside possible unreliable data on rates, and the inconsistent monitoring of early leavers, the high school dropout issue in Croatia has not been systematically researched. Existing research suggests that dropouts belong to socially vulnerable groups (Milas et al., 2010) and that dropout risk factors are low parental education level, low household income (Matković, 2010), poor academic achievement, school year repetition (Ferić et al., 2010), demotivation, negative grades, curricula perceived as difficult, and absenteeism (Strugar, 2011).

As can be observed, the existing research has mostly focused on academic and social risk factors. Our study, therefore, was designed to consider the full range of individual, family, school, and social risk factors, with a view to creating a typology of Croatian high school dropouts.

The Present Study

The present study contributes to the current body of evidence on heterogeneity of high school dropouts and does so using a non-American sample and a qualitative approach, with the aim of developing a typology that will inform psychosocial interventions for young Croatian high schoolers. Accordingly, our research questions were as follows:

Recruitment

The data presented in this article are part of a larger research project that uses a mixed-methods approach that combines quantitative data from a longitudinal study with semi-structured interviews conducted with high school dropouts. In accordance with the Croatian Ethics Code for Research with Children (Ajduković & Kolesarić, 2003), the researchers convened meetings with parents in the sampled schools to present the study and inform parents. The students filled out an informed consent form for participation in the study that also covered participation in the qualitative part of the study.

As our cohort of students (n = 1,402) was monitored over their first three high school years, information from school psychologists and pedagogues was used to identify dropouts. Each school year, these psychologists and pedagogues sent the research team a list of young people who had left school and the following information: gender, date of leaving, reasons for leaving, destination immediately after leaving, and contact details. The psychologists and pedagogues also contacted these young people to ask them whether they were willing to participate in interviews. If they agreed, a researcher telephoned them to provide more information about the study and arrange an interview.

A total of 77 high school dropouts were identified for the 3-year period of the study. Before each interview, participants received an invitation letter and description of the study and signed an informed consent form. Participants were guaranteed anonymity and confidentiality, were informed that they did not have to answer every question, and could withdraw at any time. As an incentive to participate in the study, and to maximize response, each participant was paid €15.

Participants

Purposive sampling was used to recruit 20 high school dropouts. Participants were seven young women and 13 young men aged 16 to 21 years, M (SD) = 18.4 (1.23) years. Nearly all the young people (n = 19) had been in vocational technical education (VTE; i.e., high schools that focused on practical skills that equipped students to directly enter the workforce and not the tertiary education system) and had left school 11 to 20 months, M (SD) = 13.4 (1.37) months, before being interviewed. Most (n = 16) had left school in the first year of VTE. At the time of the interview, seven had enrolled in alternative educational programs, four were employed, and nine were neither in employment nor in education/training (NEETs).

Data Collection

The semi-structured interviews, conducted between May 2017 and March 2019 at schools or in the participants’ homes and audio-recorded, were designed to obtain insights into school and life experiences, decisions to leave high school, and postschool plans and pathways. The interview guide was informed by life history (Goodson, 2001), that is, the young people were asked to talk retrospectively about the events and experiences that motivated them to quit high school. Seven broad topics were covered: perceptions of the school environment and teachers, family characteristics and relationships, peer characteristics and relationships, stressful life events, the decision to leave school, experiences after leaving school, and plans for the future. The interview was piloted with five young people (two males and three females) excluded from the data collection.

Data Analysis

Twenty interviews were transcribed verbatim (by M.P. and M.K.) and imported into NVivo 12 software (QSR International, 2019) for data analysis. The coding phase commenced using an inductive approach (Onwuegbuzie & Combs, 2010) to qualitative content analysis (Schreier, 2012), a method that “forces you to select certain key aspects of your material and to focus on those” (Schreier, 2012, p. 59) and which are known as categories. Accordingly, interview passages representing the range of experiences of dropouts were selected and categories were identified from the data to develop the coding frame. As categories were generated, these were fine-tuned, broken down into subcategories, and split into distinct ones or merged until no new categories were deemed necessary, a point that indicated data saturation (Morse, 2015). With the coding frame developed, the subcategories were reviewed to ensure that they met the four requirements of qualitative content analysis (Schreier, 2012), namely, unidimensionality, mutual exclusiveness, exhaustiveness, and saturation. In total, 15% of the data set was used to generate the coding frame that had medium complexity, that is, it entailed several categories but did not “reach down beyond one level of subcategories” (Schreier, 2012, p. 66). The final coding frame contained 108 subcategories grouped in 45 categories. Boyatzis’s (1998) guidance to working with good codes was applied to subcategories, whose meaning and application were likely to differ among coders during the coding process. Accordingly, subcategories had a clear and concise label, were given a definition, a description of when they were likely to occur in the data, some inclusion and exclusion criteria, and citation examples “to eliminate possible confusion” when looking for them in the data (Boyatzis, 1998, p. 31).

To perform inter-coder reliability, 25% of the data set that reflected the diversity of youth dropout experiences was selected for coding trial by two coders, P.K. and I.R. Both coders coded the data in two different NVivo projects which, upon completion, were merged and the “coding comparison query” was run, yielding a percentage agreement of 97%. Responses of negative agreement where the coders did not agree on which code to apply or omit (even by chance) were discussed to resolve inter-coder discrepancies.

The main coding phase consisted of coding the 20 interviews, including those used for the inter-coder reliability test. The final percentage agreement between P.K. and I.R. was 98%. To avoid double-coding and, hence, possible problems of validity, units of coding—that is, passages segmented in the data—were coded to either the category or only one of its subcategories but never to both using the thematic criterion approach (Schreier, 2012).

To answer the research question (What is the dropout pattern among high school Croatians?), cluster analysis (Hair et al., 2009), a multivariate statistical technique that uncovers structure in a data set by grouping similar objects in groups, called clusters, based on their similarity, was run. Cluster analysis is the recommended statistical technique when the purpose of a study is to generate a typology, describe what each type is made of, and contrast heterogeneity among clusters. In NVivo, cluster analysis was run using the coding patterns about the dropout experiences of participants as similarity unit and which excluded participants’ sociodemographic variables. The exclusion of sociodemographic data in the analysis was to avoid natural differential structures—such as gender—to predefine the typology.

To generate the typology, the “matrix coding query” was run where each interview was a row and the subcategories, or categories, the columns. The matrix cells showed 1 when there was coding at each pair of items and 0 when no coding occurred. Jaccard’s coefficient was used to calculate the similarity index for each pair of interviews. Of the 20 interviews, two were excluded from the cluster analysis as outliers (one was very short and the other had content similar to every type, but not similar enough in terms of the main factor of any type, yielding a low Jaccard coefficient). Our results are based, therefore, on coding pattern similarities for 18 interviews.

Type 1: The Poor Academic Achievers

The main characteristic of this group of young people is that poor academic achievement contributed predominantly to dropping out—the result of various negative school factors and circumstances. The fact that these young people carried over poor grades from elementary school often resulted in year repetition (most had to repeat the first grade of high school). All reported having difficulties in mastering the curriculum: it was hard for them to learn, they did not understand, they did not know how to complete homework assignments, and they needed help with their studies and with other school duties.

Well, everything was hard. Some of it I could learn, but English and maths never. (17)

They also reported a general lack of motivation to attend school.

I didn’t have any interest in school. It was the last of my priorities. I didn’t even, honestly, I didn’t even want to go to school at all. Like something pulled me away from school. I would rather stay at home and stare at a wall all day than go to school. (2)

In elementary school, the only activities in which they actively participated and showed interest were extracurricular sports. It is interesting to note that all the participants enrolled in a particular high school on the recommendation of a family member or a friend, rather than based on their own interests.

I wanted to enrol in a high school for professional drivers, but they [parents] told me to go for industrial mechanics. Because I would have to go to another town to go to the driving school. (10)

As an additional crucial circumstance, two of them mentioned that the school was very far from their home. Because of leaving very early in the morning and arriving home very late, they did not have much free time and were often too tired to do homework or study.

And so I would wake up at 3 am, so I could catch a train at 4 am, and at 5:15 am I would be in Varaždin [city] to catch a bus at 6 am . . . that’s 3 hours of travel . . . Because of that travelling I didn’t have any time. After school I would get back to Varaždin at 3 pm where I had to wait for the train to go home until 7:20 pm, so in that time I would go for coffee with friends, walk around the town. I would get home at 9 pm, take a shower and go to bed. (15)

Absenteeism, especially skipping classes, was high in this group, with one stating that she had started skipping class in the seventh grade of elementary school. This nonattendance often led to disciplinary measures.

The participants also mentioned that they would sometimes behave inappropriately in the class, although they viewed this behavior as minor mischief, for example, talking during class or not paying attention because of which they did not have any particular problems at school.

According to participants’ responses, their parents did not play an important role in their education; some said that their parents never asked them anything about school, whereas others said that while they did talk with their parents, the conversations were superficial: Parents would only ask about grades and be angry or yell if they were poor but never did anything specific about the issue. In some cases, parents supported their offspring’s decision to leave school.

I saw that I couldn’t do it, I didn’t want to study, I didn’t feel like doing anything for school. So I said to my mom “I’m going to get a job,” and she said, “You are going to study or you are going to work,” and I said “I’m going to get a job” and that was it. (17)

A minority of participants took part in risky behaviors associated with gaming, hanging out with peers, and alcohol consumption. In general, these young people attributed dropping out of high school to internal reasons: insufficient effort and lack of motivation resulting in poor grades and absenteeism. Most of them are now employed, whereas one has returned to the education system to finish high school.

Type 2: The Quiet Dropouts

The main characteristic of these young people is that, according to their own statements, they had internalizing problems that directly affected their performance at school.

I wasn’t satisfied with myself, and I became depressed after a while. Moreover, I don’t know, nothing made sense to me. Then, after a while, I didn’t want to go to school anymore, I couldn’t. So I wouldn’t go, and when I did go back to school, I saw that I’d missed a lot of the classes, and that it was too much for me to catch up. (13)

All these participants described signs of depression and some described somatization and anxiety symptoms.

I figured I was probably depressed when I realized that I didn’t want to live anymore. Then I thought: “This isn’t normal” . . . At first it wasn’t like that, it was just “I don’t feel so good today” and not wanting to go to school, and later I would skip classes so I could sleep in at home . . . My friends noticed that I wasn’t all right because I’d changed. Before I would be very active, run around, everything, and now I just wanted to stay home, I didn’t want to do anything. (14)

An additional circumstance was that some of the participants were bullied, which they said contributed to their internalizing problems.

I just got depressed from all the mocking . . . of the way I talked, of my weight, of the way I looked. (14)

Some of the young people also had physical health problems; for one of them, these led to nonattendance at school, but as they coincided with internalizing problems, when he recovered his physical health, he simply continued not going to school. Another participant had severe headaches, which doctors diagnosed as somatization:

I started to have headaches, and they wouldn’t stop for a week . . . and that lasted a long time. . . doctors said that they had a psychological basis. (18)

These young people, in general, had been successful in school: they got good grades, had no difficulty in mastering the curriculum, were interested, regularly did their homework, and studied and participated in extracurricular activities.

In elementary school I was pretty good at studying and I had an excellent memory. I did my homework by myself and I studied by myself [didn’t need any help]. At that time, I was really good at doing my homework, because I was always afraid that I would get a minus or an F if I didn’t. (13)

However, under the influence of internalizing problems, their academic performance and achievement deteriorated. For some participants, this started in the higher grades of elementary school, whereas for others it started in high school. Their absenteeism was aggravated once they started skipping class due to their internalizing problems, as indicated in the quotes above.

The participants indicated that their high school was their personal choice. Some described the difficulties they had in adjusting to high school and, because they found it more difficult than elementary school, they started getting low grades, which only added to the stress.

Well, it definitely was more demanding than elementary school, I even think I wasn’t prepared enough for it. I got a few good grades, then I started to get Fs, and it really shook me because I wasn’t used to it. (13)

Their parents were involved in their education, regularly attended parent–teacher meetings, talked to them about school, grades, and the importance of education and also actively supported their learning by encouraging and helping them.

When I would do math homework, I would ask my father for help. Because he knows math really well, so I would ask him for help. He would explain things to me, and I would do my homework. (5)

The participants attributed dropping out of high school to internal reasons that were indirectly or directly related to their internalizing problems, for example, dissatisfaction with school and with themselves, rumors, and the teasing of peers because of absenteeism (due to depression) as well as the depression itself:

When I started to worry a little bit more about my depression, I realized that I would never get better if I didn’t leave school for a while. (14)

These participants, who understood the subjective importance of education, eventually reenrolled in high school once their internalizing symptoms improved.

Type 3: The Maladjusted Dropouts

This group was characterized by the expression of externalizing problems that were already evident in elementary school but which intensified in high school: consumption of alcohol and psychoactive substances, gambling, vagrancy, violence, and spending time in the company of older youth with risky behavior.

I was too interested into getting into trouble in general, getting into fights. It felt like it was in my blood to get into fights . . . I always had plenty of girlfriends, and when you have a girlfriend you have to get into fights about girls. . . it was important to get bloody. And that was how you made friends, the next day people would buy me drinks at a bar and so on. (16)

They did not do any homework or study at all; their grades were poor and they typically repeated years in both elementary and high school. This was the only group that did not participate in any extracurricular activities.

I didn’t study, I didn’t even try, I wasn’t interested in school at all. (16)

This group was featured by a high level of absenteeism, especially skipping class, and the time not spent in class typically was spent on the abovementioned risky behaviors. One of them was expelled due to absenteeism.

It just stopped being interesting and I told myself that I didn’t want to be there. I knew I could not quit, that it’s compulsory to finish elementary school, but I knew I didn’t have to be in school every day. Therefore, I skipped classes, I went outside, it was fun, and so I continued. (8)

In addition to externalizing behaviors expressed out of school, these participants also displayed various behavioral problems at school, entering into conflict with teachers and also with peers whom they bullied.

I got along with other students mostly, but there were few of them that I didn’t like and so I would provoke them, insult them and so on. (8)

Parents were stated as not having been particularly involved in their education. Participants themselves avoided conversations about school, and their parents rarely asked them about school. If the parents attended parent–teacher meetings or otherwise found out about poor grades, they would advise them that school was important and that they had to study, but they did not get involved much further than that. The parents also generally failed to react to their offspring’s risky behaviors.

Well, I started to go to bars and I started to drink. Then I liked it and so I continued like that every day.

[Did your parents see you under the influence of alcohol?]

Well, yes, but I managed to get away with it. I said that I was tired or something . . . a couple of times they figured me out. (8)

In response to the externalizing problems, the schools often intervened, involving the police and social services; one of participants was placed in an institution for problematic youngsters.

I had conversations with social workers, they would come to our house. I would promise them anything they wanted, they would believe me, but I wouldn’t do any of what they said. So, in November they placed me in an institution. They were saying from October that they would do that, but I thought that they were only trying to scare me. (11)

The reasons for dropping out of high school were viewed as internal: insufficient engagement, absenteeism, and their own behavior at school.

I didn’t go to school, I lied saying that I didn’t have any classes, I would say that I didn’t feel like going to school. They [educators] tried to make me go to school, but I continued to behave as I wanted. And that went on until they signed me out of school. (11)

One of them recently returned to education, whereas two other participants are now a part of the NEET population.

Type 4: The Stressed Dropouts

The main characteristic of this group was that these young people experienced highly stressful life events that represented key turning points for them when attending high school. The events included disrupted family relations, discrimination based on ethnicity, and a false accusation of theft.

Some problems arose, the fights began, I was shocked. Between them [parents] it was no longer what it was before. I would fight with them and ask why it was like this? It was a terrible blow for me, and then I started to give up in school. (1)

I always knew that the Roma people were discriminated against.

[Do you think you didn’t get the work placement necessary to pass the class because you’re a Roma?]

Well, it’s obvious. Other students got them [placements], but not me. (20)

While some were more and others less interested in school, all of them used to do their homework and study and all of them used to get reasonable grades, until the life-changing event occurred and led to negative changes, including a loss of interest in school, deteriorating grades, repeated years, behavioral problems, fights with teachers, bullying by peers, consumption of marijuana and alcohol, skipping class, and feelings of helplessness, hopelessness, and depression.

Note that participants did not see such problems as the specific reason for dropping out; rather, those issues were just symptoms that could be related back to a specific stressful event or obstacle to successful continuation of education. For example, a girl who was falsely accused of theft had other issues, but she recognizes that the false accusation was the final straw—She could not bear the situation itself or its consequences for her.

For four years I had been suffering from depression and emotional disorders . . . because I was emotionally unstable, and then all of this [a false accusation of theft], and so they [peers] used to say things about me, and I snapped, and then I really don’t know, I just couldn’t stand it. (19)

The parents of this type of student were involved and showed interest in their offspring’s education, supported them, talked to them about school, and tried to help them deal with their school issues.

My dad always asked me every day after school “how was school, what did you do today?” Therefore, I would tell him what happened in school every day. It was great for me. Because I felt important, like, that they wanted to know what I did all day. (20)

As in the previous group, one of the participants returned to education to finish high school, whereas the other four continue to be NEETs.

Implications for Practice and Future Directions

A more comprehensive understanding of the different types of students who drop out of high school will enable the design and implementation of preventive programs in schools. Such interventions, adapted to the needs, vulnerabilities, and strengths of each group, would modify developmental and educational pathways and promote academic and social adjustments. Findings from this study for Croatia may be important for the creation of targeted interventions, the modification of school policies, and the development of early warning indicators, all designed to keep young people enrolled in school (e.g., Freeman & Simonsen, 2015; Legters & Balfanz, 2010; McDermott et al., 2018). Our findings also suggest the need for a personalized multicomponent approach that considers unique risk profiles and protective factors for individuals.