Predicting School Dropout Research Articles

Effectiveness of Artificial Intelligence Models for Predicting School Dropout: A Meta-Analysis

School dropout is a major concern in the educational systems of all countries. In recent years, artificial intelligence is playing an important role in predicting school dropout in the different educational stages of formal education. In this context, it is crucial to know that these predictions are accurate and understandable. This meta-analytic study aims to investigate the effectiveness of dropout prediction models conducted until May 2022. The databases used are Web of Science, Scopus, PubMed, ERIC, PsyInfo, Dialnet and Scielo. 15 studies with a sample size of 199,015 participants are analyzed. The meta-analysis uses a random-effects proportions model with 95% confidence interval. Statistical evidence indicates that artificial intelligence models performed well (91%; 95% CI = 89-93%) in predicting dropout; specifically, the Decision Tree model significantly (95.3%; 95% CI = 93-98%) predicts dropout better than other models such as Random Forest, Artificial Neural Network, Support Vector Machines, Logistic Regression and Stacking Ensemble. Consequently, more models should be applied in the dropout field with larger numbers of participants to confirm these findings and improve the quality of education.

Poorer self-reported mental health and general health among first year upper secondary school students do not predict school dropout: a five-year prospective study.

Education is important for socioeconomic, work and health status; thus, dropping out of secondary school is of major concern. In Norway, 1 out of 5 is dropping out from upper secondary education. Academic performance is a known predictor for dropout, but the role of mental and general health status is studied less. By use of student data collected during the first school year we examined the accumulated risk of school dropout over 5 years. Students entering upper secondary school in a North-Norwegian region (Troms County) completed a comprehensive questionnaire during August 2010 (N = 1,676, 69% response rate). The contribution of mental and general health problems in predicting five-year dropout was of primary interest, adjusted for demographics and academic performance. One-third of the students had dropped out after 5 years. A logistic regression analysis showed no significant effect of mental and general health problems on dropout. Among the covariates, higher grades from lower secondary education reduced the chance of dropping out (OR = 0.31; p < 0.001). Subgroup analyses showed that students in the vocational track reported poorer mental and general health, compared to students in the general track, but this difference was not related to dropout. General track students were also less likely to drop out than vocational track students (OR for dropout 0.66; p < 0.05). In conclusion, lower grades from lower secondary education represented a warning flag for school dropout during upper secondary education whereas mental health issues were not.

Advancing school dropout early warning systems: the IAFREE relational model for identifying at-risk students.

There is a global effort to address the school dropout phenomenon. The urgency to act on it comes from the harmful evidence that school dropout has on societal and individual levels. Early Warning Systems (EWS) for school dropout at-risk student identification have been developed to anticipate and help schools have a better chance of acting on it. However, several studies point to a doubt that Correct EWS may come too late because they use only publicly available and general student and school information. We hypothesize that having a tool to assess more subjective and inter-relational factors would help anticipate where and when to act to prevent school dropout. This study aimed to develop a multidimensional measure for assessing relational factors for predicting school dropout (SD) risk in the Brazilian context. We performed several procedures, including (a) the specialized literature review, (b) the item development of the Relational Factors for the Risk of School Dropout Scale (IAFREE in Portuguese), (c) the content validity analysis, (d) a pilot study, and (e) the administration of the IAFREE to a large Brazilian sample of high school and middle school students (N = 15,924). After the theoretical steps, we found content validity for five relational dimensions for SD (Student-School, Student-School Professionals, Student-Family, Student-Community, and Student-Student) that include 12 facets of risk factors. At the empirical stage, confirmatory analysis corroborated the proposed theoretical model with 12 first-order risk factors and 5 s-order dimensions (36 items). Further, through the Item Response Theory analysis, we assessed the individual item parameters of the items, providing a brief measure without losing psychometric quality (IAFREE-12). We discuss how this model may fill gaps in Correct EWS models and how to advance it. The IAFREE is a good measure for scholars investigating the risk of SD. These results are important for implementing an early warning system for SD that looks into the complexity of the school dropout phenomenon.

Longitudinal associations between positive attributes and psychopathology and their interactive effects on educational outcomes.

Psychopathology is associated with impaired learning and early termination of schooling, whereas positive attributes are associated with better educational outcomes. However, it is important to understand if and how psychopathology and positive attributes longitudinally impact each other so we could shed light on where to intervene to promote educational outcomes through these constructs. A large prospective school-based community cohort of youths (5-15years of age, 45% female) were assessed and followed up for 3years (n = 2010; 80% retention). We assessed the longitudinal impact of positive attributes (Youth Strength Inventory) and psychopathology (bifactor model of Strengths and Difficulties Questionnaire) using a cross-lagged panel model. We also used generalized mixed effects models to investigate how these both constructs predict school dropout and literacy, adjusting for confounders and testing their interaction. Positive attributes negatively predicted, and were negatively predicted by, the general factor of psychopathology and conduct problems in the cross-lagged panel model. Positive attributes (OR = 0.57, 95% CI [0.44, 0.73], p < 0.001) and specific conduct symptoms (OR = 2.33, 95% CI [1.64, 3.33], p < 0.001) predicted school dropout, whereas the general factor of psychopathology predicted lower literacy ability (β = - 0.08, 95% CI [- 0.11, - 0.05], p < 0.001). However, the protective association of positive attributes on school dropout decreases as the general factor of psychopathology increases. These findings provide new evidence that positive attributes and psychopathology mutually influence each other over development and have interactive effects on educational outcomes.

Décrochage scolaire : la relation élève-enseignants peut-elle l'emporter contre le quotient intellectuel?

Le Québec détient le plus haut taux de non-diplomation au Canada. Les conséquences individuelles et sociales sont nombreuses, ayant des répercussions à très long terme. Bien qu’un faible quotient intellectuel (QI) soit un facteur de risque important du décrochage scolaire, certains facteurs pourraient moduler cette relation. L'objectif de l'étude est d'examiner l'effet modérateur de la relation élève-enseignants sur le lien entre le QI et le décrochage scolaire à l’aide des données provenant de l'étude Stratégie d'Intervention Agir Autrement (SIAA). Selon les analyses de régression logistique, le score de QI contribue à prédire le décrochage scolaire. Cependant, la relation élève-enseignants ne contribue pas au modèle de prédiction et ne modère pas l'association entre le QI et le décrochage scolaire.

Statistical Learning for Predicting School Dropout in Elementary Education: A Comparative Study

School dropout is a significant challenge for the education system. This phenomenon is present in different environments, modalities, and stages of education. In the Brazilian scenario, despite advances in some respects as a reduction of indexes, combating evasion is still one of the significant efforts. Identifying the factors that involve school dropout is supported by different decision support techniques such as Statistical Learning. Statistical learning consists of a method set for exploring and understanding data to establish an association between explanatory and response variables and develop an accurate model. We propose to examine the use of some regression methods commonly used in the Statistical Learning literature for estimating school dropout in the context of elementary school from the state of Pernambuco. The data involves educational indicators, and we defined phases in the study to understand, prepare, and model the data. For prediction, we apply models for estimating school dropout using kernel-based and linear regression methods. We measured the performance by the prediction error from the test data set using Mean Absolute Error and Root Mean Square Error. We considered Statistical tests to confirm the results. The findings show that kernel-based models are effective alternatives to provide greater precision in the estimation of school dropout in scope studied. The reason to explore more accurate predictive models is supporting intervening and targeting the most at-risk students of scholar dropout. The study provides knowledge about the applied scenario supporting policies to mitigate the problem.

A Combined Model based on Clustering and Regression to Predicting School Dropout in Higher Education Institution

A Combined Model based on Clustering and Regression to Predicting School Dropout in Higher Education Institution

Predictive Models for Imbalanced Data: A School Dropout Perspective

Predicting school dropout rates is an important issue for the smooth execution of an educational system. This problem is solved by classifying students into two classes using educational activities related statistical datasets. One of the classes must identify the students who have the tendency to persist. The other class must identify the students who have the tendency to dropout. This problem often encounters a phenomenon that masks out the obtained results. This study delves into this phenomenon and provides a reliable educational data mining technique that accurately predicts the dropout rates. In particular, the three data classifying techniques, namely, decision tree, neural networks and Balanced Bagging, are used. The performances of these classifies are tested with and without the use of a downsample, SMOTE and ADASYN data balancing. It is found that among other parameters geometric mean and UAR provides reliable results while predicting the dropout rates using Balanced Bagging classifying techniques.

“Big Data” in Educational Administration: An Application for Predicting School Dropout Risk

Purpose: In an era of unprecedented student measurement and emphasis on data-driven educational decision making, the full potential for using data to target resources to students has yet to be realized. This study explores the utility of machine-learning techniques with large-scale administrative data to identify student dropout risk. Research Methods: Using longitudinal student records data from the North Carolina Department of Public Instruction, this article assesses modern prediction techniques, with a focus on tree-based classification methods and support vector machines. These methods incorporate 74 predictors measures from Grades 3 through 8, including academic achievement, behavioral indicators, and socioeconomic and demographic characteristics. Findings: Two of the assessed classification algorithms predict high school graduation and dropping out correctly for more than 90% of an out-of-sample student cohort. Findings reveal a shift toward lower dropout incidence in regions hit hardest by the economic recession of 2008, especially for male students. Implications for Research and Practice: Machine-learning procedures, as demonstrated in this study, offer promise for allowing administrators to reliably identify students at risk of dropping out of school so as to provide targeted, intensive programs at the lowest possible cost.

Predicting school dropout with administrative data: new evidence from Guatemala and Honduras

ABSTRACTSchool dropout is a growing concern across Latin America because of its negative social and economic consequences. Identifying who is likely to drop out, and therefore could be targeted for interventions, is a well-studied prediction problem in countries with strong administrative data. In this paper, we use new data in Guatemala and Honduras to estimate some of the first dropout prediction models for lower-middle income countries. These models correctly identify 80% of sixth grade students who will drop out within the next year, performing better than other commonly used targeting approaches and as well as models used in the U.S.

Poverty moderates the association between gender and school dropout in South African adolescents

ABSTRACTThis study examined prospective associations between poverty, gender, and school dropout in a large community sample of South African adolescents (baseline: n = 3515, follow-up: n = 3401, 57% female, age range at baseline: 10–17 years, mean age at baseline = 13.45). School dropout was defined as being enrolled in school at baseline assessment but no longer enrolled in school at follow-up assessment. Poverty was measured at baseline assessment using an index of access to the eight highest socially perceived necessities for South African children and adolescents. Demographic characteristics including child gender and age, province, and urban versus rural location were recorded at baseline assessment and controlled for in the analysis. As predicted, higher poverty scores (AOR = 2.01, p < .001) were associated with increased odds of school dropout 1 year later. Gender was not a significant predictor of school dropout (AOR = 1.56, p = .07) but did interact with poverty (AOR = 0.66, p = .04) in predicting school dropout. However, our initial hypothesis that the impact of poverty on school dropout would be stronger for girls than boys was not supported. Instead, results indicated that while girls were at elevated risk of school dropout at low and mean levels of poverty, at high levels of poverty this gender difference was no longer evident. Findings suggest that vulnerable boys should not be neglected in policies to improve retention in education in contexts of extreme poverty.

Why Wait? Early Determinants of School Dropout in Preventive Pediatric Primary Care.

BackgroundTo answer the question of what bio-psychosocial determinants in infancy, early and middle childhood, and adolescence predict school drop-out in young adulthood, we approached the complex process towards school dropout as a multidimensional, life-course phenomenon. The aim is to find signs of heightened risks of school dropout as early as possible which will eventually help public health workers in reducing these risks.MethodsIn a case-control design, we used data from both the Preventive Pediatric Primary Care (PPPC) files (that contain information from birth onwards) and additional questionnaires filled out by 529 youngsters, aged 18–23 years, and living in the South-east of the Netherlands. We first conducted univariate logistic regression analyses with school-dropout as the dependent variable. Backward and forward stepwise analyses with the significant variables were done with variables pertaining to the 0 to 4 year period. Remaining significant variables were forced into the next model and subsequently variables pertaining to respectively the 4 to 8, 8 to 12 and 12 to 16 year period were introduced in a stepwise analysis. All analyses were cross-validated in an exploratory and confirmatory random half of the sample.ResultsOne parent families and families with a non-Western background less often attended the health examinations of the PPPC and such less attendance was related to school dropout. The birth of a sibling (OR 0.63, 95% CI 0.43–0.93) in infancy and self-efficacy (OR 0.53, 95% CI 0.38–0.74) in adolescence decreased the odds of school dropout; externalizing behavior (OR 2.81, 95% CI 1.53–5.14) in middle childhood and (sickness) absence (OR 5.62, 95% CI 2.18–14.52) in adolescence increased the risks.ConclusionTo prevent school dropout, PPPC professionals should not wait until imminent dropout, but should identify and tackle risk factors as early as possible and actively approach youngsters who withdraw from public health care.

The Reciprocal Links Between School Engagement, Youth Problem Behaviors, and School Dropout During Adolescence

Drawing on the self-system model, this study conceptualized school engagement as a multidimensional construct, including behavioral, emotional, and cognitive engagement, and examined whether changes in the three types of school engagement related to changes in problem behaviors from 7th through 11th grades (approximately ages 12-17). In addition, a transactional model of reciprocal relations between school engagement and problem behaviors was tested to predict school dropout. Data were collected on 1,272 youth from an ethnically and economically diverse county (58% African American, 36% European American; 51% females). Results indicated that adolescents who had declines in behavioral and emotional engagement with school tended to have increased delinquency and substance use over time. There were bidirectional associations between behavioral and emotional engagement in school and youth problem behaviors over time. Finally, lower behavioral and emotional engagement and greater problem behaviors predicted greater likelihood of dropping out of school.

Grade Retention and Seventh-Grade Depression Symptoms in the Course of School Dropout among High-Risk Adolescents

The relationship between grade retention and adolescent depression in the course of school dropout is poorly understood. Improving knowledge of the mechanisms involving these variables would shed light on at-risk youth development. This study examines whether depression in adolescence moderates the relationship between grade retention and school dropout in a high-risk sample. Seventh-grade students (n = 453) from two low-SES secondary schools in Montreal (Quebec, Canada) were followed from 2000 to 2006. Self-reported lifetime and seventh-grade depression were measured with the Inventory to Diagnose Depression. Primary school grade retention, and secondary school dropout status was obtained through the Ministry of Education of Quebec registries. Sixteen percent of participants reported lifetime depression, and 13% reported depression in seventh-grade. Nearly one third (32%) of the sample dropped out of school. Logistic regression models were used to estimate moderation effects predicting school dropout six years later. Findings indicated that students with grade retention were 5.54 times more likely to drop out of school. Depression in seventh grade increased by 2.75 times the likelihood of school dropout. The probability of dropping out for adolescents combining both grade retention and seventh-grade depression was 7.26 times higher than it was for those reporting grade retention only. The moderating effect of depression was similar for boys and girls. Depression is a significant vulnerability factor of low educational attainment aggravating the risk associated with grade retention. Experiencing depression at the beginning of secondary school can interfere with school perseverance particularly for students who experienced early academic failure.

Predicting School Dropout and Adolescent Sexual Behavior in Offspring of Depressed and Nondepressed Mothers

To examine predictors of school dropout and adolescent sexual behavior in offspring of depressed and nondepressed mothers. Possible moderators of the relation between maternal depression and these outcomes also were explored. Participants were 240 mothers and adolescents assessed annually from 6th through 12th grade. Interviews and questionnaires measured the chronicity and severity of the mother's depressive episodes, the mother's educational attainment, socioeconomic status, the presence of a father, the adolescent's IQ, externalizing behaviors, and substance use disorders. Substance use disorders before ninth grade significantly predicted higher rates of both school dropout and adolescent sexual behavior. Lower levels of mother's educational attainment and higher rates of adolescent's externalizing behaviors in grades 6 through 8 predicted higher school dropout. Higher IQ was associated with a lower likelihood of dropping out among offspring of never or moderately depressed mothers, but not for offspring of chronic/severely depressed mothers. Among offspring of never or moderately depressed mothers, the presence of a male head of household was associated with lower rates of adolescent sexual behavior, but not among offspring of chronically/severely depressed mothers. These findings have implications for the development of programs aimed at preventing behavior problems in high-risk adolescents.

Trajectories of Prosocial Behavior and Physical Aggression in Middle Childhood: Links to Adolescent School Dropout and Physical Violence

Trajectories of prosocial behavior and physical aggression between 6 and 12 years of age were identified for a sample (N=1,025) of males. The trajectories were then used to predict school dropout and physical violence at age 17. Using a group‐based semi‐parametric method, two trajectories of prosociality (low and moderate declining) and three trajectories of physical aggression (low, moderate, and high declining) were obtained. Only a small minority (3.4%) of the boys were characterized by both high aggression and moderate prosociality. Physical aggression predicted both school dropout and physical violence, but contrary to expectations, prosocial behavior did not have additive or protective effects.

School Dropout as Predicted by Peer Rejection and Antisocial Behavior

This longitudinal study assessed the prediction of school dropout from measures of antisocial behavior, social preference, and achievement. The sample was assessed at 8th grade (n= 516) and was expanded at 10th grade (n= 1157). In the 8th, but not 10th grade, rejected–antisocial students had higher dropout rates than those in other groups. Seperate logistic regression analyses were then conducted using the 8th‐ and 10th‐grade data. Across both sets of analyses, aggression and achievement, but not social preference, predicted subsequent dropout. These results, in conjunction with those of others, suggests that social preference does not uniquely predict school dropout. The possibility exists, however, that youth who are both antisocial and rejected may be at heightened risk for school dropout.