Academic failure of students attending high schools can be considered a serious issue of concern for countries. In fact, they can lead to dropout which has severe consequences both at the individual and social level. In order to prevent this phenomenon, it is important to understand its causes. This thesis aims at investigating which are the determinants of academic failure in terms of grade retention for students attending the first year of upper secondary school, which is demonstrated to be the most critical, in the province of Piacenza (northern Italy). Data about 2071 students who attended the first year of high school during academic year 2018/2019 were used for this research. Such data were collected from upper secondary schools, lower secondary schools and the INVALSI institution. Logistic regression, classification tree and random forest techniques were adopted to carry out analyses. Furthermore, hierarchical models for each of these methods were implemented basing random effects on two different grouping criteria: high school and class. Propensity Score Matching was used to investigate the effects of particular treatments on the probability of failing the ninth grade of education. On one hand, findings of this research confirmed what other scholars already discovered: core subjects’ grades are relevant determinants as well as prior academic performance. On the other hand, they added some important information. For instance, not following the orientation advice and being ambitious when choosing the high school track was found to increase by at least 11% the risk of being held back to repeat grade 9.
L’insuccesso degli studenti durante la scuola secondaria di secondo grado è un problema serio che preoccupa molti paesi. Infatti, può presto trasformarsi in abbandono scolastico e avere conseguenze negative non solo per l’individuo ma anche per l’intera società. Perciò, è importante agire cercando di prevenire questo fenomeno, iniziando dal comprenderne le cause. L’obiettivo di questa tesi è investigare quali sono le determinanti dell’insuccesso scolastico inteso come bocciatura alla fine del primo anno di scuola superiore, che è stato dimostrato essere il più critico. Per questo scopo sono stati utilizzati dati provenienti dalle scuole secondarie e da INVALSI riguardanti 2071 studenti che hanno frequentato il primo anno di scuola superiore nell’intera provincia di Piacenza durante l’anno scolastico 2018/2019. I metodi utilizzati in questa ricerca sono stati: regressione logistica, alberi di classificazione e foreste casuali. Inoltre, per ognuno dei metodi appena descritti, sono stati implementati modelli multilivello basando gli effetti casuali su due diversi criteri di raggruppamento: la scuola superiore e la classe. Infine, la tecnica del Propensity Score Matching è stata utilizzata per analizzare gli effetti di particolari trattamenti sulla probabilità di essere bocciati. I risultati di questa ricerca confermano quello che già altri autori avevano evidenziato, ovvero che i voti delle materie principali sono predittori importanti così come la precedente carriera scolastica. Inoltre, si è scoperto che non seguire il consiglio orientativo ed essere ambiziosi scegliendo una scuola secondaria di secondo grado che è comunemente ritenuta più impegnativa aumenta di almeno l’11% la probabilità di ripetere il primo anno.
Early warning systems to predict and prevent academic failure in high school
Baldi, Veronica;BOLZONI, FILIPPO
2019/2020
Abstract
Academic failure of students attending high schools can be considered a serious issue of concern for countries. In fact, they can lead to dropout which has severe consequences both at the individual and social level. In order to prevent this phenomenon, it is important to understand its causes. This thesis aims at investigating which are the determinants of academic failure in terms of grade retention for students attending the first year of upper secondary school, which is demonstrated to be the most critical, in the province of Piacenza (northern Italy). Data about 2071 students who attended the first year of high school during academic year 2018/2019 were used for this research. Such data were collected from upper secondary schools, lower secondary schools and the INVALSI institution. Logistic regression, classification tree and random forest techniques were adopted to carry out analyses. Furthermore, hierarchical models for each of these methods were implemented basing random effects on two different grouping criteria: high school and class. Propensity Score Matching was used to investigate the effects of particular treatments on the probability of failing the ninth grade of education. On one hand, findings of this research confirmed what other scholars already discovered: core subjects’ grades are relevant determinants as well as prior academic performance. On the other hand, they added some important information. For instance, not following the orientation advice and being ambitious when choosing the high school track was found to increase by at least 11% the risk of being held back to repeat grade 9.| File | Dimensione | Formato | |
|---|---|---|---|
|
2020_10_Baldi_Bolzoni.pdf
accessibile in internet solo dagli utenti autorizzati
Dimensione
4.1 MB
Formato
Adobe PDF
|
4.1 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/166297