INTRODUCTION

Truancy, disruptive behaviour, family and personal difficulties, failing grades and repeating grades (repeated failure) are the prelude to early school leaving (hereafter referred to as ESL). This is a progressive disengagement process. It is not something that happens suddenly (Robin & Burger, 2020). Examining the above-mentioned concepts, the term school failure is repeatedly the subject of discussion. For Fernández Enguita et al. (2010), failure has a denotative and connotative value. The denotative is due to the fact that there is no clear definition of it, as for some it would consist of not completing Compulsory Secondary Education (ESO), and for others, not completing post-compulsory secondary education, while all forms of failure, repetition or delay could be included; that is, partial failures can mark a difficult path to success. For example, in terms of school year repeaters in Andalusia, who account for 33.3%, a value well above the average for OECD countries with 11.4% (PISA, 2019). The connotative would entail the disqualification and even stigmatisation of the pupil, blaming them exclusively, with the consequent de-responsibilisation of the institution. This failure is sometimes due to serious absenteeism processes. This refers to the regular and continuous non-attendance of pupils in basic and compulsory education at the school they attend, without illness or major cause to justify their absence. According to data from the PISA Report 2015 (OECD, 2016), the percentage of students who say they have missed at least one day of school without justification is 25% in Spain, compared to 19% on average in OECD and EU countries. Failing at school, failing exams and sometimes not being allowed to progress to the following year, and missing classes, are all factors that contribute to the process of “disengagement”, which often ends up with students dropping out of school, as reported in previous studies (González-Losada et al., 2015). The terms “disengagement”, “dropping out” or disaffection have been widely used in studies in the English-speaking sphere. This is evidenced by the plethora of scientific articles that use the terms. Specifically, according to González and Bernárdez-Gómez (2019), there are students who become disengaged and end up dropping out. This phenomenon is well documented and has a complex and multifaceted nature, in which multiple contexts (school, classroom, families, community) and people (pupils, teachers, classmates, family members) interrelate, and in which factors of a very different kind (personal, social, cultural, economic, academic, school, etc.) intervene. These authors build on previous work such as that by Patton and Price (2010), in which they distinguish between students who are “visibly disengaged” (disruptive, apathetic and disengaged) and those who are disengaged, albeit not obviously (students who are adept at achieving high grades, despite lacking interest and aspiration). In similar terms, Sodha and Guglielmi (2009) distinguish between active disengagement (visible and manifest in low achievement, negative attitudes, misbehaviour or avoidance of participation) and passive disengagement (students who passively withdraw from their education, distancing themselves cognitively and emotionally by being less visible). Notable among these works is the study by Fredricks et al. (2019) in which they review school commitment/engagement.

Having clarified these differences between concepts, we need to address ESL. According to the recommendations of the Council of the European Union of 28 July 2009 (Council of the European Union, 2011) on policies to reduce early school leaving, they specify that: “The term ‘early school leavers’ is used in relation to people who leave education and training with only lower secondary education. …” (p. 1). The dropout rate is calculated from data provided by the Economically Active Population Survey (EAPS) of the National Statistics Institute (INE), taking as a reference the International Standard Classification of Education (ISCED), ensuring comparability with results from other countries (INE, 2015). Thus, according to the EAPS, the percentage of 18-24 year olds whose highest level of education corresponds to one of the ISCED levels 0 (Pre-school), 1 (Primary), 2 (Compulsory Secondary) or 3 (Post-compulsory), short cycle and who are not studying or training in the four weeks prior to the survey, are dropouts (thus ruling out levels 3.A, Baccalaureate; 3.B, Intermediate Level Training Cycles; and 3.C, Occupational Training, which pursues specialisation in work skills.) (INE, 2015). The same definition can also be found in Eurostat (2020). One of the priority objectives of the Europe 2020 strategy is that the value of this indicator should not exceed 10%. Spain, in 2019, presented an ESL rate of 17.3% (21.4% men and 13% women), far exceeding the rates of other European countries. In this sense, “the level of early school leaving remained at very high levels (around 30%) between 2000 and 2008, without any improvement until 2009, when the indicator began to decrease” (Bayón-Calvo, 2019, p.50). In the present study, the dropout rate is 21.6% on average (16.9% for women and 26.1% for men), Ministry of Education and Vocational Training, 2020a). Given the magnitude of the problem, within the Sustainable Development indicators, the issue is monitored. Specifically, the 2030 Agenda for Sustainable Development (INE, 2021) includes the goal of ensuring inclusive and equitable quality education and promoting lifelong learning opportunities for all, and the corresponding indicator (4.1.2. Completion rate for primary and secondary education).

In parallel, the results of studies such as the Trends in International Mathematics and Science Study (TIMSS), which assesses the mathematical and scientific competence of pupils in Primary Education (PE) and year 2 of Compulsory Secondary Education (ESO), or those of the Programme for International Student Assessment (PISA), carried out by the Organisation for Economic Co-operation and Development (OECD), highlight the poorer performance of Spanish students compared to the average for OECD countries (OECD, 2018, 2019).

All these circumstances have serious individual and social consequences (Lessard et al., 2010). For example, these students have less access to the labour market, lower salaries and a higher risk of exclusion (Latif et al., 2015; UNESCO, 2015). This leads to high costs, a high rate of unskilled population and lower productivity, given the low level of education. This vulnerability, victimisation and exclusion in young early leavers is higher than in other age groups, according to international studies (Escudero et al., 2013; European Union, 2013; Sañudo Guerra, 2022).

In this context, it is necessary to analyse Spanish secondary schools, as they are the object of our study and the educational stage where the risk of dropout is highest. Previous research has identified issues that strongly influence the risk of ESL: teachers abuse traditional methodologies ( Rué, 2006); students do not recognise the usefulness of their studies and tend to get bored in their classes (Moeller et al., 2020); they think that marks or grades are not important (González-Losada et al., 2015; Tarabini y Curran, 2015); pupils do not feel included and this conditions their behaviour in terms of school regulations (Strayhorn, 2019; Johnson et al., 2020); nor do they recognise the centre as a place of their own and problems are identified in relations between peers (González-Falcón et al., 2016); the degree of commitment is low this conditions their performance (Fredricks et al., 2019) and relationships with teaching staff (Martin & Collie, 2019; Skinner et al., 2008).

Janosz et al. (1997) identified how the main predictors of school dropout in the last quarter of the 20th century, i.e. school, family, behavioural, social and personality variables, can be grouped into three main factors: personal, socio-family and school. These conclusions are reaffirmed in subsequent studies, most of which are also quantitative, and others based on a qualitative methodology (Aristamuño, 2009; Calero, 2006; Janosz et al., 2008; Lessard et al., 2008; Suberviola, 2021). In the personal scope, research identifies skills, needs, interests and motivations as influencing the risk of dropout. Lower school commitment and participation are directly related to a higher likelihood of dropping out. Montecinos (2018) emphasised the lack of school participation and the absence of alternative routes to further studies. McGrath and Van Bergen (2015), focusing on pupil behaviour and temperament, they claim that students who are disruptive, aggressive or antisocial in class are at high risk of dropping out.

Janosz et al. (2008), considered that low engagement and unstable school records point to a high risk of dropping out and often coincide with personal, family or social problems. Gubbels et al. (2019), in turn, identified risk factors related to the characteristics of young people such as age (the older they are, the higher the risk of absenteeism and consequently of failure and dropping out).

The second group of risk factors is linked to the family: low socioeconomic levels (Andalusia has one of the highest unemployment rates in the national scope -INE, 2022-), psychosocial factors of the family and family-educational relationships (Suberviola, 2021), the attitude of parents towards studies, the type of language and communication existing between them; the promotion of reading, future expectations or the value they place on education (Janosz et al. 2008; Marchesi, 2003; Vitaro et al., 2001). The family can be expected to support the optimal development of children, given its educational influence as an agent of socialisation (Torío López, 2004 and Fajardo Bullón, et al., 2017). Peña et al. (2016) stated in their study that parental lack of concern is the most important element in school dropout. In the case of poorly motivated and committed pupils, their parents further decrease their involvement in their education, and their teachers find it more difficult to establish positive relationships with them. These problems are due, among other variables, to the paths that the student has followed. The involvement of families and the backing and support of the school institution are therefore essential when students are at risk. Hence the importance of establishing connections between family and school (Álvarez and Martínez, 2016).

Other studies highlight how increasing parents’ educational attainment decreases the likelihood of dropout (Cerruti & Binstock, 2004; Binstock & Cerruti, 2005). Research such as that of Battin-Pearson et al. (2000) or Fortin et al. (2004) concludes that children who perceive little cohesion, conflicts, communication problems, lack of organisation in the family or who come from broken families, are more at risk of dropping out than other pupils Gubbels et al. (2019) extended these problems to the characteristics of the parents by referring to psychological problems or difficulties at work.

The third group of factors are social and school aspects. They are usually focused on the role of teachers (Janosz, et al., 2008; Lessard et al., 2008, 2010, Sancho & Esteban, 2007). Some of the causes that most clearly intervene in student performance are also identified, such as the average socioeconomic level of the school, or repetition in previous years. This study by Gubbels et al. (2019) identified the features of their schools as determining factors (very high ratios, high repetition rates, low teacher quality); or the characteristics of their peer group (antisocial behaviour, truancy, delinquency, etc.).

In view of the above, and given the high dropout and unemployment rates in the Andalusian community, this study was carried out in order to examined in depth some of the risk factors for dropout. In this paper, we analyse the socio-familial characteristics and their influence on the behaviours and attitudes of Andalusian secondary school pupils, presenting the results of the validation of two scales of an ad hoc instrument.

METHOD

The research presented is quantitative and hypothetical-deductive, following a survey method and cross-sectional design. Although the research was more extensive (two questionnaires were administered: one for teachers and one for pupils). The overall aim of this work is to confirm the influence of socio-familial characteristics on students’ attitudes and behaviour in the classroom. As specific objectives, we propose:

• To analyse secondary school students’ perceptions of their socio-familial characteristics, attitudes and behaviour in the classroom.
• To assess the construct validity of the “socio-familial characteristics” and “students’ attitudes and behaviour in the classroom” scales.
• To confirm the influence of socio-familial characteristics on students’ attitudes and behaviour in the classroom.

The initial hypothesis was formulated as follows: socio-familial characteristics influence students’ attitudes and behaviour in the classroom.

Sample

The sample used focuses on the educational community of 3rd and 4th year of Compulsory Secondary Education (ESO) in Andalusia. The sample consists of a total of 1426 pupils (N: 160.000; p<0,01; sampling error: + 4%) from 16 challenging ESO schools. The selection of the schools was intentional and carried out by the Andalusian Ministry of Education. The selection criteria were absenteeism and failure rates above the average for Andalusian schools and high rates of sanctions for disruptive behaviour. This made it easier for researchers to access and identify the schools, as these data are not publicly available.

Regarding the sample characteristics, 94.7% came from state schools (1350) compared to 5.3% (75) from schools, i.e. centres whose administration is mainly public but which are privately run. The selected sample were in 3rd (52.5%) and 4th (47.5%) year of ESO. As for gender, 49.2% were men and 50.5% were women, thus achieving a balanced representation in both groups. The mean age was 14.98 years, with a standard deviation (SD) of 0.96.

T-tests were carried out to study the differences in the sample distribution, with the following results: in terms of the educational level of parents, in general, no significant differences were found between the type of studies of parents (p=.38). However, there was a tendency for mothers’ educational attainment to be higher in line with university education (Figure 1). This is not the case for tertiary education/ vocational training, where there is a higher presence of fathers than mothers.

With regard to employment status, as shown in Figure 2, there are significant differences between the employment status of fathers and mothers (P≤.001), with a higher percentage of fathers in paid work (76.5%) than mothers (48.3%), with mothers devoting more of their time to housework (41.5%) than fathers (0.6%). There is also a higher percentage of unemployed fathers (15.5%) than mothers (7.5%).

Finally, in reference to some student performance indicators, 58.9% (840 pupils) had never repeated a year, 27% had repeated one year (385 pupils), 12.3% had repeated two years (176 pupils), 0.9% had repeated three years (13 pupils) and 0.1% had repeated four years (1 pupil). The average number of subjects failed in the previous academic year was 1.27 (SD= 2.08), with considerable variability in the responses to this item. As for the type of subject repeated (Table 1), the highest percentages are accumulated in Mathematics (23%), English (18.9%), Spanish Language and Literature (18%), Biology and Geology (15.5%), Social Sciences (15.2%), Physics and Chemistry (10.5%), Technology (9.7%) and French (6.5%). The lowest percentages were concentrated around the subjects of Plastic and Visual Arts (3.2%), Civic Ethics (3%), Music and Physical Education, with 2.6% respectively, and Computer Science (0.8%).

Instrument

When designing our instrument, we took other scales already validated into account. We reviewed the work of Fortin et al. (2006), who studied risk factors around three main dimensions: personal, family and school, with pupils in 7th grade (equivalent to 1st year of ESO); the work of Lessard et al. (2008; 2010), focused on school factors, with pupils who had dropped out of school. The work by Janosz et al. (2000), which analysed school, family and peer relationship factors to predict dropout risk; Vitaro et al. (2001), who used questionnaires to analyse family factors; and those of Archambault et al. (2009a, 2009b), studying the relationship between pupil participation and engagement with the school institution in secondary schools. All these research studies served as our references when gathering information on the incidence of risk factors. What they all show is the complexity of the problem of dropping out and the heterogeneity of the individuals at risk, as well as the different weight that each of the factors has in the school leaving process.

In our study, two scales were applied: one measuring pupils’ perceptions of socio-familial characteristics (Annex I) and the other gauging their perceptions of pupils’ attitudes and behaviour in the classroom (Annex II).

The scales were evaluated by a group of experts who assessed the relevance of the content and clarity of the wording of the items and made some changes to the wording of the items. Subsequently, a pilot study was carried out in a secondary school in the city of Huelva, which serves a socially disadvantaged population and deploys different actions to ensure access, promotion and permanence of these students in the education system. The outcomes obtained in terms of the construct validity of the scales “socio-family characteristics” and “student attitudes and behaviour in the classroom” are shown in the results section.

The socio-familial characteristics scale consists of 12 items, obtaining a Cronbach’s alpha of .52, a low initial value but acceptable in the early stages of exploratory studies. (Nunnally, 1987). However, the items were reviewed through the corrected total correlation, and no item was identified whose removal increased Cronbach’s alpha significantly compared to that obtained. The pupil classroom behaviour scale reached an alpha of .78 for 20 items. Each of these items is measured on a Likert-type scale from 1 to 5 (1 being never and 5 always).

Procedure

Given that the Education Administration (General Delegation of Education of the Andalusian Regional Government, the highest authorities responsible for this area in this community), carried out the selection of the schools with the criteria specified above and notified them of their participation in the project. The procedure for application of the instrument was simple. We contacted them and arranged the appointments with the centres, as the surveys were applied in person. The questionnaires were administered on paper. The research team travelled to the different Andalusian provinces and schools selected. The questionnaires were administered and collected on the same day, going through the different classrooms. This ensured anonymity and data protection.

Prior to implementation, two meetings were held with the General Delegation of Education in order to obtain ethical approval for the process. At the first meeting, the project, its objectives and the procedure to be applied were presented. The instrument was left to be studied by the Delegation’s Ethical Committee. After this and at the second meeting, we were given permission to apply the procedure, as well as the list of selected schools.

For the data analysis, an exploratory factor analysis of each variable was carried out using the two scales proposed in this study, with the aim of analysing the students’ perception of the socio-familial characteristics and the attitudes and behaviour of pupils in the classroom. The means and standard deviations were obtained for each of the variables, and homogeneity was studied by means of the corrected item-total correlation, allowing us to analyse whether the items are related to each other, i.e. they measure the same variable, as well as their ability to discriminate between the items. The response options for the Likert-type items were grouped into three levels: low (1-2), medium (3) and high (4-5), in order to obtain their percentages.

To gather evidence of the construct validity of the “socio-familial characteristics” and “student attitudes and behaviour in the classroom” scales, an exploratory study was carried out, using the maximum likelihood estimation method and Varimax rotation to analyse the factor structure of each scale. The internal consistency of the complete scale and the resulting subscales of the factor analysis was analysed using Cronbach’s alpha.

All analyses were performed with the SPSS data statistics software, version 17. Finally, to confirm the influence of socio-familial characteristics on pupils’ attitudes and behaviour in the classroom, structural equations were used through the AMOS v.18 programme, including in the model the factors extracted in the exploratory analysis, adopting as a reference the criteria established by Byrne (2010) and Kline (2010) (CMIN/DF between 2 and 5, CFI and IFI > .9, RMSEA < .06 and HOELTER > 200).

RESULTS

The results are presented in response to the study objectives.

Objective 1. To analyse students’ perceptions of socio-familial characteristics and pupils’ attitudes and behaviour in the classroom

Tables 2 and 3 show the items that make up the socio-familial characteristics and student attitudes and behaviour in the classroom scales for each of the items. Mean and standard deviations, the homogeneity through the corrected item-total correlation and the response percentages are analysed according to the three levels of grouping: low (1-2), medium (3) and high (4-5).

There is sufficient variability for all items, as the minimum and maximum available values were reached for all items. The results obtained are described below (Tables 2 and 3).

With regard to socio-family characteristics, the main ones perceived are those related to the importance of education in the family environment, such as: “my family thinks that studying is very important” (95.9%), “my family are concerned about my grades” (91.7%), “my family is confident of my ability to succeed in my studies” (85.9%), “my family speaks well of the teachers and school” (77.3%), “I have enough time to study” (64.4%).

In contrast, there is little presence of those items related to lack of family involvement, such as “my family goes to the school to talk to the teachers” (37.1%) or “my family helps me with my homework assignments” (20.1%); nor is there a perceived presence of those characteristics of the social environment, such as “many of my friends are thinking about giving up studying” (18.2%) or “in my neighbourhood there are bad relations among young people” (8.2%).

As for the homogeneity index, corrected item-total correlation was positive for all items, after inverting the items: D3, D11, D12 and D10, so that they contribute in the same sense as the measurement by the questionnaire. We found items whose values are lower than <0.2, D6, D10 and D12, and decided to rule out these items, following the theoretical recommendations (Frías-Navarro, 2019).

In terms of classroom attitudes and behaviour, there is a lower proportion of disruptive acts in class. All the items listed in this time mentioned received very low scores: “They throw me out of the class” (1.4%); “I pester my classmates” (1.4%); “I deliberately disturb the class” (2.3%); “I usually skip class” (3.8%); “I feel isolated” (4.8%); “I come late to class” (5.4%).

Pupils scored high on those items that highlight the value of education (B13, B11, B14, B9, B2), in all of which more than 50% scored high (4-5).

On the other hand, different opinions are observed with regard to positive attitudes towards education; “I like studying” (low: 47.7%; average: 23.6%; high: 28.7%); “I like the school” (low: 51.3%; average: 21.5%; high: 27.2%).

The corrected item-total correlation is positive for all items. To this end, the items were reversed: B1, B10, B16, B3, B20, B4, B7, B17, B6, B5, B15, thus contributing in the same direction in the measurement. The decision was taken to remove item B15, whose value was less than <0.2 (Frías-Navarro, 2019).

Objective 2. To assess the construct validity of the “socio-familial characteristics” and “students’ attitudes and behaviour in the classroom” scales

For the factor analysis, items that presented difficulties in the previous analysis were removed and items with factor loadings below .30 were eliminated, as these items are considered to discriminate weakly (Nunnally y Bernstein, 1995). The analyses were relevant given the high KMO indices of the scales, ranging from .707 to .842, which allows for the convenience of performing the factor analysis. On the other hand, Bartlett’s test of sphericity shows the applicability of the factor analysis of the scales (significance index <.001). The resulting factor structure is presented in table 4.

For the “socio-familial characteristics” scale, the results yielded a total of 3 factors explaining 51.77% of the variance. A first factor refers to the “importance” that the family gives to their children’s education, explaining 25.11% of the variance and consisting of four items (D7, D8, D9, D2). A second factor, “involvement”, consists of two items (D5, D4) that refer to the family’s willingness to help in the educational tasks of their children, as well as to attending school to talk to teachers, explaining 15.05% of the variance. The third factor refers to “social” aspects and accounts for 11.59% of the variance (items D11 and D12).

Cronbach’s alpha for the full scale is .53 (8 items). The alpha for each of the extracted factors is .53 for F1, .47 for F2 and .45 for F3. These alpha values, although lower, are considered sufficient in the early stages of exploratory studies ( Nunnally, 1987). Two less consistent items are identified (D11 and D12), whose removal increases the alpha by .6 for 6 items. Nevertheless, we decided to keep them in the study, as their correlations with the rest of the items are significant (p< .001) and they yield factor loadings above .5, both of which are good indicators of construct validity. On the other hand, we want to study how these two items related to a type of socio-familial characteristic (social aspects) relate to students’ attitudes and behaviour in the classroom.

For the student attitudes and behaviour in the classroom scale, the factor analysis yielded a total of 4 factors that explain 48.27% of the variance.

Factor 1 consists of items B2, B10, B13, B11, and B8, accounting for 24.40% of the total variance and measuring the value of education as perceived by the student body. Factor 2 is made up of those items that measure acts of disruption in the classroom (B4, B7 and B5) and explain 8.96% of the total variance. Factor 3 explains 7.78% of the variance, saturating items related to having a positive attitude towards education (B12 and B18).

Factor 4 cannot be explained with the underlying dimension property; to be a dimension it must be expressed in several ways, not with a single item.

Regarding Cronbach’s alpha, an index of .64 (10 items) was obtained for the full scale. The following indices were found for each of the factors: F1: .75 (5 items), F2: .65 (3 items) and F3: .72 (2 items).

Objective 3. To confirm the influence of socio-familial characteristics on students’ attitudes and behaviour in the classroom

The structural equation model depicted in Figure 3 reflects the three types of socio-familial characteristics perceived by students (social, importance of education in the family, family involvement) and their influence on student behaviours in the classroom (disruptive acts, value of education, positive attitude towards education, etc.). In this sense, the model represents the correlations between the three types of perceived socio-familial characteristics and regressions or influences of these on students’ attitudes and behaviours in the classroom.

Initially, two models were explored (table 6). Model 1, which considered the relationship of influence between socio-familial characteristics and student attitudes and behaviour in the classroom. Model 2 (Figure 3) improved the fit indices by establishing an influential relationship between the value of education on disruption and the value of education on positive attitude.

The goodness-of-fit indices indicated that the data fitted the model discretely well. On the other hand, the high rates of variance explained in the factors of disruption (47%), value of education (49%) and positive attitude towards education (20%), as well as the regression and correlation indices, support the conformity of the model (Figure 1).

Table 7 shows the model estimates identifying the relationships explored with standardised and non-standardised values.

First of all, in relation to the correlations between the perceived socio-familial characteristics, a high degree of correlation was found between “Importance of education” and “Involvement” (r = 0.41, p< 0.001). Conversely, between “Importance of education” and “Social characteristics” (r = -0.37, p< 0.001); and, to a lesser extent, between “Social characteristics” and “Involvement” (r = 0.15, p= 0.262).

With regard to the regression analysis between factors, the following results can be observed in the model. On the one hand, the perceived importance that the family gives to education has both direct and indirect and significant effects on most of the pupils’ attitudes and behaviours in the classroom. Social characteristics are less influential on acts of disruption in the classroom. Family involvement does not have a clear influence on students’ attitudes and behaviour in the classroom, and the regression indices are not significant in any of the cases.

On the other hand, the importance given by the family to education as perceived by students seems to exert a direct and significant influence on the “value of education” (β =0.63, p<0.001). This in turn has a direct and significant influence on the positive attitude of pupils (β = 0.49, p<0.001) and an inverse influence on acts of disruption in the classroom (β = -0.48, p<0.001). In other words, the more importance students perceive the family as placing on education, the more value they place on it, generating positive attitudes towards education and decreasing disruptive acts in the classroom. With regard to acts of disruption in the classroom, the direct effect of social problems perceived by pupils is confirmed (β= 0.37, p<0.001).

DISCUSSION AND CONCLUSIONS

The main aim of this study was to confirm the influence of socio-family characteristics on the attitudes and behaviour of Andalusian pupils in the classroom as determinants of early school leaving. In this sense, the main results obtained point to the importance of education within the family as the main characteristic. These findings are in line with what is reflected in the studies of Janosz et al. (2008), Marchesi (2003), Suberviola (2021) and Vitaro et al. (2001), where it is identified that the value given to education by families is a determining factor regarding the risk of dropping out of school, while highlighting the role of the family context as the primary socialisation group ( Torío López, 2004). The main socio-familial characteristics include social characteristics, the importance of education in the family environment and family involvement. In this sense, the results of this study, in response to the first objective of this work (to analyse students’ perceptions of socio-family characteristics, attitudes and behaviour in the classroom), show that families are concerned about and attach a positive value to studies. That is, their children perceive academic support from their parents, although supervision, monitoring and providing resources and support for their learning (in terms of Reschly & Christenson, 2019) seem not to be perceived by the students.

As for the pupils, a large group values education and disruptive behaviours are rare, which are good indicators of school success; although, on the other hand, the diversity of opinions regarding the fact that they do not like studying or school, a result that coincides with other research in the same context (González-Losada, et al., 2015), means that we should keep an eye on these risk factors. Although there is a correlation between inappropriate behaviour and the risk of students dropping out of school, this study shows, as mentioned, a very low proportion of disruptive behaviour or misbehaviour in the classroom: expulsion [81.4%], being a nuisance [1.4%], arriving late to class [5.4%], according to the students’ perception of the risk of dropping out of school. This may be due to the protective factor of families. As for the second objective, to assess the construct validity of the “socio-familial characteristics” and “pupils’ attitudes and behaviour in the classroom” scales, the factorial structure was corroborated, with 3 factors (the importance that the family attaches to the education of their children, the family’s involvement in helping their children and attendance at the school to talk to teachers and social aspects) in the “socio-familial characteristics” scale. On the scale of pupils’ attitudes and behaviour in the classroom, the factor analysis projected 3 factors: value of education as perceived by students; disruptive acts in the classroom; and having a positive attitude towards education.

With reference to the third objective, to confirm the influence of socio-familial characteristics on students’ attitudes and behaviour in the classroom, we highlight the evidence that socio-familial characteristics do influence their attitudes and behaviour in the classroom, both directly and indirectly. Directly, because social problems (such as poor relationships between young people and wanting to drop out of school) predict acts of disruption in the classroom. In agreement with Torío López (2004) and Suberviola (2021), the importance that the family attaches to education predicts that students will give value to education. And indirectly, because the value of education, predisposed by the importance given to it by the family, has a moderating effect on acts of disruption. In other words, the more value students place on education, the less disruptive acts seem to occur in the classroom. The value of education also has a moderating effect on positive student attitudes.

The model outlines the proximity of the three socio-familial characteristics through the strength and significance of the correlation. Thus, there is a gradual increase in the distance between the different characteristics, in such a way that the importance that the family attaches to education is closer to family involvement, and family involvement, at the same time, is further away from social problems, and in the same sense, the importance that the family attaches to education is closer to the absence of social problems.

Therefore, a certain arrangement is shown between the different socio-family characteristics that show a close nature to each other. On the one hand, involvement seems to be associated with importance and, on the other hand, an increase in importance seems to decrease social issues. In this sense, involvement, in the equation, does not appear as a predictor of students’ attitudes and behaviour in the classroom, but it does appear to be associated with importance.

As limitations, we first highlight some of the model’s fit values (CFI=.87; IFI=.87; NFI= .85; TLI= .81) that are below the minimum and therefore susceptible to improvement. However, these indices are measured in a range of 0-1 and are considered acceptable values when they are close to or above .9, being those that show the best fit of the model. In our case, although we are at the limits of the recommended values, we think it is important to propose an explanatory model that serves, albeit in an exploratory way, to gain a more precise understanding of the phenomenon in question.

We want to understand the model beyond the isolated interpretation of fit indices, as it allows us to examine the established relationships, which, although they do not achieve a good fit, do present initial acceptance values within the established limits, and also fit the theoretical foundations reviewed.

In second place, another limitation is based on the need to contrast the students’ opinions with those of the teachers and families, in order to help us to adjust the objectives set. On the other hand, the use of the questionnaire provides us with quantitative results, so it would be interesting to consider new research with other qualitative instruments to complement and triangulate the results, as proposed in studies such as those by Janosz et al. (2000) or González-Losada et al. (2015), among others. In addition, we must remember the bias of the students’ answers due to “social desirability”, and that the sample selection was intentional (defined by the educational authorities of the Andalusian community). Both issues have conditioned the results and will be taken into account in future research.

Finally, in line with the contributions of Hernández Prados and Alcaraz Rodríguez (2018), Janosz (1997) and Suberviola (2021), the present study has allowed us to gain more in-depth knowledge of the risk factors for school dropout related to socio-family characteristics, student attitudes and behaviours in the classroom, and pupils’ behaviour in the classroom. Along these lines, we propose that schools should tackle school dropout from the earliest educational stages and take care of transitions between educational stages in order to guarantee the commitment of all students to their studies (Fredricks et al., 2019). These interventions can be approached from various perspectives: on the one hand, they should promote the design and development of educational projects of a preventive nature that focus on furthering specific content related to personal development (self-esteem, self-knowledge, self-concept, self-control, social skills, life skills, resilience, decision-making, etc. and teaching-learning strategies (intellectual work techniques, motivation, learning to learn, active learning methodologies, etc.). These actions should be reflected in the guidance, action and tutorial plans of educational centres, in which families should be involved so that the proposals are developed in a coordinated manner within and outside the educational context.

Socioeconomic characteristics in Andalusia are complex and have consequences on students’ engagement, motivation and academic improvement. This is a variable that is difficult to tackle. However, in order to alleviate these realities, it is necessary that schools, together with families, define and combine ways to work on positive expectations and support the educational process both academically and motivationally. Family influence, which plays an important role in school success, has been studied and counteracts the bad influences of peers (Reschly & Christenson, 2019). In this sense, teachers should receive training to improve these strategies for working with families, as they have difficulties and gaps in this area of intervention (Epstein & Sanders, 2006).

Another fundamental line of work to be developed should focus on the demotivation and low commitment of students. In this sense, in the Andalusian community we find different programmes and actions. Among them, the so-called Performance and Learning Improvement Programmes (PMAR) approved in the Organic Law for the Improvement of Educational Quality, although Márquez and Indarramendi (2022) note that these programmes are having little effect as a strategy to prevent school failure. And the Second Chance Schools, “... an educational resource that has adopted flexible organisational structures and proactive procedures, based on individualised attention, dialogue and the involvement of agents from outside the school itself” (Prieto, 2015, p.110).

In short, we agree with Bayón-Calvo (2019) that Spain, and especially Andalusia, has a high rate of early school dropout that needs to be reduced. Also, it is necessary to continue to analyse in depth the factors that affect it and the evaluation of the programmes that are implemented to prevent it. Along with this, it is important to emphasise the complexity of the problem, given the multiple factors involved and the need to mobilise global education policies in order to provide more comprehensive responses to this troubling situation.

Acknowledgments

We would like to thank our colleague Sebastián González-Losada (now retired) for having always counted on us in his projects. Funding: Project PRY 102/11 “Pupils at risk of dropout: Analysis of contextual factors in the Autonomous Community of Andalusia”. Principal Investigator: Dr. Sebastián González-Losada. University of Huelva. Funded by the Andalusian Studies Centre.

REFERENCES

Álvarez, L., & Martínez, R. A. (2016). Cooperación entre las familias y los centros escolares como medida preventiva del fracaso y del riesgo de abandono escolar en adolescentes. Revista Latinoamericana de Educación Inclusiva, 10(1), 175-191. http://dx.doi.org/10.4067/S0718-73782016000100009

Archambault, I., Janosz, M., Fulli, J., & Pagani, L. (2009a). Student engagement and its relationship with early high school dropout. Journal of School Health, 3(32), 651-670. https://doi.org/10.1016/j.adolescence.2008.06.007

Archambault, I., Janosz, M., Morizont, J., & Pagani, L. (2009b). Adolescent behavioral, affective, and cognitive engagement in school: Relationship to dropout. Journal of School Health, 9(79), 408-415. https://doi.org/10.1111/j.1746-1561.2009.00428.x

Aristamuño, A. (2009). El abandono de los estudios del nivel medio en Uruguay: Un problema complejo y persistente. Revista Iberoamericana sobre Calidad, Eficacia y Cambio en Educación, 7(4), 182-197. https://cutt.ly/Q88CQP2

Battin-Pearson, S., Newcomb, M. D., Abbott, R. D., Hill, K. G., Catalano, R. F., & Hawkins, J. D. (2000). Predictors of early high school dropout: A test of five theories. Journal of Educational Psychology, 92, 568–582. https://doi.org/10.1037/0022-0663.92.3.568

Bayón-Calvo, S. (2019). Una radiografía del abandono escolar temprano en España: Algunas claves para la política educativa en los inicios del siglo XXI. Revista Complutense de Educación, 30(1), 35-53. https://doi.org/10.5209/RCED.55465

Binstock, G., & Cerrutti, M. (2005). Pathways to social exclusion: High school dropout in Argentina. Comparative Perspectives on Adolescent Development in a Globalizing World. [Ponencia de congreso]. Standford University Center for Advanced Study in the Behavioral Sciences (CASBS), Stanford.

Byrne, B. (2010). Structural equation modelling with AMOS (2ª ed.). Taylor and Francis Group.

Calero, J. (2006). Desigualdades tras la educación obligatoria: Nuevas evidencias. Documento de trabajo de la Fundación Alternativas, n. 83/06.

Cerrutti, M., & Binkstock, G. (2004). Camino a la exclusión: Determinantes del abandono escolar en el nivel medio en la Argentina. [Ponencia de congreso]. Primer Congreso de la Asociación Latinoamericana de Demografía, Caxambú, Brasil.

Consejo de la Unión Europea. (2011). Recomendación del Consejo de 28 de junio de 2011 relativa a las políticas para reducir el abandono escolar prematuro (Texto pertinente a efectos del EEE) 2011/C 191/01. http://bitly.ws/sPYQ

Epstein, J. L., & Sanders, M. G. (2006). Prospects for change: Preparing educators for schools, family, and community partnerships. Pedagogy Journal of Education, 81, 81-120. https://doi.org/10.1207/S15327930pje8102_5

Escudero, J. M., González, M. T., Moreno, M., Nieto, J. M., & Portela, A. (2013). Estudiantes en riesgo, centros escolares de riesgo: Respuestas educativas al alumnado en situaciones de vulnerabilidad. DM.

European Union. (2013). Study on educational support for newly arrived migrant children. Publication Office of the European Union.

Eurostat. (2020). Early leavers from education and training by sex. https:// ec.europa.eu/eurostat/statistics-explained/index.php?title=Early_leavers_from_education_and_training#A_higher_proportion_of_early_leavers_are_ men_.E2.80.93_analysis_by_sex

Fajardo Bullón, F., Maestre Campos, M., Felipe Castaño, E., León del Barco, B., & Polo del Río, M. I. (2017). Análisis del rendimiento académico de los alumnos de Educación Secundaria Obligatoria según las variables familiares. Educación XX1, 20(1), 209-232. https://doi.org/10.5944/educxx1.17509

Fernández Enguita, M., Mena Martínez, L., & Riviere Gómez, J. (2010). Fracaso y abandono escolar en España. Fundación La Caixa.

Frías-Navarro, D. (2019). Apuntes de consistencia interna de las puntuaciones de un instrumento de medida. Universidad de Valencia. https://cutt.ly/B889zX8

Fortin, L., Royer, E., Potvin, P., Marcotte, D., & Yergeau, E. (2004). La prédiction du risque de décrochage scolaire au secondaire: Facteurs personnels, familiaux et scolaires. Revue Canadienne des Sciences du Comportement, 36(3), 219-231. https://doi.org/10.1037/h0087232

Fortin, L., Marcotte, D., Potvin, P., Poyer, E., & Joly, J. (2006). Typology of students at risk of dropping out of school: Description by personal, family and school factors. European Journal of Psychology of Education, 4(21), 363-383. https://doi.org/10.1007/BF03173508

Fredricks, J. A., Resschly, A. L., & Christeson, S. L. (Eds.) (2019). Handbook of student engamennt interventions. Working with disengaged students. Elsevier.

González, T., & Bernárdez-Gómez, A. (2019). Elementos y aspectos del centro escolar y su relación con la desafección de los estudiantes. Revista de Investigación en Educación, 17(1), 5-19. https://cutt.ly/A889EOw

González-Falcón, I., Coronel, J. M., & Correa, R. I. (2016). School counsellors and cultural diversity management in Spanish secondary schools: Role of relations with other educators and intervention model used in care of immigrant students. Intercultural Education, 27(6), 615-628. https://doi.org/10.1080/14675986.2016.1259462

González-Losada, S., García-Rodríguez, M. P., Ruiz, F., & Muñoz, J. M. (2015). Factores de riesgo de abandono escolar desde la perspectiva del profesorado de Educación Secundaria Obligatoria en Andalucía (España). Profesorado. Revista de Curriculum y Formación del Profesorado 19(3), 226-245.

Gubbels, J., Van der Put, C. E., & Assink, M. (2019). Risk factors for school absenteeism and dropout: A meta-analytic review. Journal of Youth and Adolescence 48, 1637–1667. https://doi.org/10.1007/s10964-019-01072-5

Hernández Prados, M. A., & Alcaraz Rodríguez, M. (2018). Factores incidentes en el abandono escolar prematuro. Revista de Investigación en Educación, 16(2), 182-195. https://cutt.ly/I88NIsu

Instituto Nacional de Estadística. (2015). Abandono temprano de la educación-formación. http://bitly.ws/sPZF

Instituto Nacional de Estadística. (2021). Indicadores de la Agenda 2030 para el Desarrollo Sostenible. https://cutt.ly/S88NMN7

Instituto Nacional de Estadística. (2022). Tasas de paro por distintos grupos de edad, sexo y comunidad autónoma. https://cutt.ly/288N3Ri

Janosz, M., LeBlanc, M., Boulerice, B., & Tremblay, R. E. (1997). Disentangling the weight of school dropout predictors: A test on two longitudinal samples. Journal of Youth and Adolescence, 26(6), 733-762. https://doi.org/10.1023/A:1022300826371

Janosz, M., LeBlanc, M., Boulerice, B., & Tremblay, R. (2000). Predicting different types of school dropouts: A typological approach with two longitudinal samples. Journal of Educational Psychology, 92(1), 171-190. https://doi.org/10.1037/0022-0663.92.1.171

Janosz, M., Archambault, I., Morizot, J., & Pagani, L. (2008). School engagement trajectories and their differential predictive relations to dropout. Journal of Social Issues, 64(1), 21-40. https://doi.org/10.1111/j.1540-4560.2008.00546.x

Johnson, R. M., Strayhorn, T. L., & Parler, B. (2020). “I just want to be a regular kid:” A qualitative study of sense of belonging among high school youth in foster care. Children and Youth Services Review, 111(104832), 1-8. https://doi.org/10.1016/j.childyouth.2020.104832

Kline, R. (2010). Principles and practice of structural equation modelling (3ª ed.). The Guilford Press.

Latif, A., Choudhary, A., & Hammayun, A. A. (2015). Economic effects of student dropouts: A comparative study. Journal of Global Economics, 3(2), 2-5. http://dx.doi.org/10.4172/2375-4389.1000137

Lessard, A., Butler-Kisber, L., Fortin, L. Marcotte, D., Potvin, P., & Royer, E. (2008). Shades of disengagement: High school dropouts speak out. Social Psychology of Education, 11, 25-42. https://doi.org/10.1007/s11218-007-9033-z

Lessard, A., Poirier, M., & Fortin, L. (2010). Student-teacher relationship: A protective factor against school dropout? Procedia Social and Behavioral Sciences, 2, 1636-1643. https://doi.org/10.1016/j.sbspro.2010.03.250

Marchesi, A. (2003). El fracaso escolar en España. Fundación Alternativas.

Márquez, C., & Indarramendi, C. (2022). La prevención del fracaso escolar en educación secundaria desde el Programa de Mejora del Rendimiento y del Aprendizaje. Revista Internacional de Educación para la Justicia Social, 11(1), 157-173. https://doi.org/10.15366/riejs2022.11.1.009

Martin, A. J., & Collie, R. J. (2019). Teacher–student relationships and students’ engagement in High school: Does the number of negative and positive relationships with teachers matter? Journal of Educational Psychology, 111(5), 861–876. https://doi.org/10.1037/edu0000317

McGrath, K. F., & Van Bergen, P. (2015). Who, when, why and to what end? Students at risk of negative student–teacher relationships and their outcomes. Educational Research Review, 14, 1-17. https://doi.org/10.1016/j.edurev.2014.12.001

Ministerio de Educación y Formación Profesional. (2019). PISA 2018 Programa para la Evaluación Internacional de los Estudiantes Informe español. Secretaría General Técnica.

Ministerio de Educación y Formación Profesional. (2020a). Sistema estatal de indicadores de la educación. Secretaría General Técnica. http://bitly.ws/sPZL

Ministerio de Educación y Formación Profesional. (2020b). TIMSS 2019. Estudio Internacional de Tendencias en Matemáticas y Ciencias Informe español. Versión preliminar.

Moeller, J., Brackett, M. A., Ivcevic, Z., & White, A. E. (2020). High school students’ feelings: Discoveries from a large national survey and an experience sampling study. Learning and Instruction, 66(101301), 1-15. https://doi.org/10.1016/j.learninstruc.2019.101301

Montecinos, C., Castro, G., Díaz, R., Manríquez, L., & Edwards, A. (2018). Trabajo interprofesional en los centros escolares: Cambiando trayectorias de vulnerabilidad a la deserción en estudiantes secundarios. RELIEVE, 24(2), Artículo M4. http://doi.org/10.7203/relieve.24.2.13386

Nunnally, J. (1987). Teoría psicométrica. Trillas.

Organización de las Naciones Unidas para la Educación, la Ciencia y la Cultura. (2015). A growing number of children and adolescents are out of School as aid fails to meet the mark. UNESCO Institute for Statistics and the Education for All Global Monitoring Report, July, 1–13. https://cutt.ly/h882eFY

Organización para la Cooperación y el Desarrollo Económicos. (2016). PISA 2015 Results (Volume II): Policies and Practices for Sucessful Schools. OECD Publishing https://cutt.ly/n881cIE

Organización para la Cooperación y el Desarrollo Económicos. (2019). PISA 2018 Results (Volume I): What Students Know and Can Do. OECD Publishing. https://doi.org/10.1787/5f07c754-en

Organización para la Cooperación y el Desarrollo Económicos. (2018). Programa para la evaluación Internacional de Estudiantes. Resultados PISA 2018. https://cutt.ly/Q881Rg5

Patton, A., & Price, D. (2010). Learning Futures: The Engaging School, principles and practices. https://cutt.ly/8881ImZ

Peña, J. C., Soto, V., & Calderón, U. (2016). La influencia de la familia en la deserción escolar: Estudio de caso en estudiantes de secundaria de dos instituciones de las comunas de Padre las Casas y Villarrica, Región de la Araucanía, Chile. Revista Mexicana de Investigación Educativa, 21(70), 881-899.

Prieto, B. (2015). El camino desde la vulnerabilidad escolar hacia el desenganche educativo. El papel de las escuelas de segunda oportunidad en la estrategia contra el abandono educativo. Profesorado. Revista de Currículum y Formación del Profesorado, 19(3), 110-125.

Reschly, A. L., & Christenson, S. L. (2019). The intersection of students engagement and families: A critical connection for achievement and life outcomes. En J. A., Fredricks, A. L. Resschly, & S. L. Christeson (Eds.), Handbook of Student Engamennt Interventions. Working with disengaged students. Elsevier.

Robin, S., & Burger, K. (2020). Negative life events, self-efficacy, and social support: Risk and protective factors for school dropout intentions and dropout. Journal of Educational Psychology, 112(5), 973–986. http://dx.doi.org/10.1037/edu0000406

Rué, J. (Coord.) (2006). Disfrutar o sufrir la escolaridad obligatoria. ¿Quién es quién ante las oportunidades escolares? Octaedro.

Sancho, M. A., & De Esteban, M. (Eds.) (2007). Estudio sobre el abandono prematuro de la escuela en les Illes Balears: diagnóstico y tipología de los grupos de Riesgo. Documento de trabajo. Fundación Europea Sociedad y Educación.

Sañudo Guerra, L. (2022). Del abandono a la permanencia escolar en secundaria. Profesorado. Revista de Currículum y Formación de Profesorado, 26(1), 213-233.

Skinner, E., Furrer, C., Marchand, G., & Kindermann, T. (2008). Engagement and disaffection in the classroom: part of a larger motivational dynamic? Journal of Educational Psychology, 100(4), 765-781. https://doi.org/10.1037/a0012840

Sodha, S., & Guglielmi, S. (2009). A stitch in time: Tackling educational disengagement. Interim report. Demos.

Strayhorn, T. L. (2019). College students’ sense of belonging: A key to educational success for all students (2ª ed). Routledge.

Suberviola, I. (2021). Análisis de los factores predictivos del abandono escolar temprano. Vivat Academia. Revista de Comunicación, 154, 25-52. http://doi.org/10.15178/va.2021.154.e1373

Tarabini, A., & Curran, M. (2015). El efecto de la clase social en las decisiones educativas: Un análisis de las oportunidades, creencias y deseos educativos de los jóvenes. Revista de Investigación en Educación, 13(1), 7-26.

Torío López, S. (2004). Familia, escuela y sociedad. Aula Abierta, 83, 35-52. http:// hdl.handle.net/10651/26953

Vitaro, F., Larocque, D., Janosz, M., & Tremblay, R. (2001). Negative social experiences and dropping out of school. Educational Psychology: An International Journal of Experimental Educational Psychology, 4(21), 401-415. https://doi.org/10.1080/01443410120090795

---

Received: 25 February 2022
Accepted: 02 March 2023
Published: 13 June 2023