Articles

Publish at November 08 2023 Updated November 08 2023

Reduce potential dropouts

Can AI detect at-risk learners?

Young man's support group

School systems differ from country to country, yet all fear one thing: dropping out. From secondary schools to the highest levels of education, dropouts are often perceived as failures by organizations. The phenomenon also worries the general public, and parents in particular, who dread the prospect of their child dropping out. A survey of parents in France showed that 51% had this fear.

So the general aim is to prevent learners dropping out at all costs. To achieve this, however, we need to be able to identify early on those at risk of dropping out of school, a task that is no easy task given the multiple factors involved.

Risk factors

The most common signs of dropping out are repeated lateness, absenteeism and falling grades. However, even knowing this doesn't explain why a pupil or student is about to throw in the towel.

We need to look at the underlying reasons, which are manifold. In the case of young adults, the cost of living, personal events and, above all, educational difficulties are among the reasons why many of them want to give up. The feeling of finding oneself alone with all these problems accentuates frustration and discouragement. That's why it's so important for faculties to provide a wide range of psychosocial and tutoring tools to support them.

The question of accompaniment and inclusion plays a key role, including in adult education. This study, carried out in Chile, shows that schools have a major responsibility in providing services to ensure that everyone feels included and encouraged in the process, especially those trying to get back on track after dropping out. The researchers argue that the South American country should do more to encourage these programs by offering more state-paid opportunities for some to complete their schooling.

This is a reminder that social inequalities play a part in the reasons for dropping out. Parents without a high level of education will unconsciously or unconsciously lead their children to drop out of school more quickly if they feel out of place.

A study carried out ten years after pupils entered sixth form in 2007 showed that 38% of drop-outs came from homes where the parents were unemployed, had only a school-leaving certificate or no qualifications at all, 22% from service worker families, 19% from unskilled worker families and 4% from the children of executives or teachers.

In a context where studies are not considered important, it's hard for students to motivate themselves to go to school every day, especially if they have learning difficulties. They will compare themselves to their peers, realize that they don't understand as quickly, become demoralized and no longer want to experience what is perceived as humiliation every day.

This is all the more true if the student is subjected to harassment at school or does not have a quality friendship with a classmate. Research carried out in Quebec in 2020 has shown the importance of these links with peers in school motivation, especially in adolescence. The social context even plays a role in student engagement. For example, in Japan, with its strict school culture and lack of support, a certain percentage of young people refuse to go to school. The only way for them to get back in has been through alternative routes, including schools where e-sports are part of the weekly activities, in addition to classes.

Artificial intelligence to predict future dropouts?

Many countries are turning to the technology of the moment: artificial intelligence. Algorithms are able to rapidly analyze the multiple data sets of each student, noting the warning signs.

Schools in Uruguay, for example, have begun to use this type of technology to flag up repeated absences among certain learners. The software also examines all the factors that could contribute to a potential drop-out: students' family responsibilities, socio-economic background, discriminatory attitudes or bullying, etc. An approach of this kind is also being studied in Egypt, with a view to creating software that would be able to note all the risks and target students to better support them in staying in school.

Of course, it would be interesting to be able to rely on artificial intelligence to immediately detect people at risk and intervene. The problem at the moment is the arbitrary nature of the data. Certainly, absenteeism is a sign, but will the algorithm be able to understand whether it is caused by illness or a serious family situation? Even this article, which claims that this technology holds great promise for universities, admits that certain personal criteria are difficult to integrate into the analysis. For example, a student who only posts a few messages on a course platform may appear "unmotivated", even though he or she may be shy or have less need to communicate as part of the learning process.

Will AI be able to make such distinctions? Doubts remain. Last but not least, the whole ethical issue is raised by these uses, including data protection and the constant surveillance effect.

A learner targeted by an algorithm could feel that he or she is under constant surveillance by the institution, and no longer trust the teaching staff and other adults. Not to mention that artificial intelligence alone is not infallible, and to avoid mistakes, human judgment will be needed to discern real dropout potential and not raise false alarms.

Photo: en.depositphotos.com

References:

"AI and big data deployed to prevent school dropout." ITU Hub. last updated November 1, 2022. https://www.itu.int/hub/2022/11/predictive-ai-school-dropout-prevention-latin-america/.

"Friendship and high school dropout risk: mediating effect of school motivation." CTREQ - RIRE. Last updated November 30, 2022. https://rire.ctreq.qc.ca/amitie-et-risque-de-decrochage-scolaire-au-secondaire-effet-mediateur-de-la-motivation-scolaire/.

Argence, Joséphine. "How to help a student experiencing a school dropout?" PARENTS.fr. Last updated April 18, 2023. https://www.parents.fr/enfant/ecole/vie-scolaire/comment-aider-un-eleve-vivant-un-decrochage-scolaire-1022450.

Chiaramonte, Francesco. "Preventing dropout: how AI is shaping the future of education." Francesco Chiaramonte. last updated March 31, 2023. https://fchiaramonte.com/ai-to-prevent-dropout/.

Contreras-Villalobos, Tabata, Verónica López, Enrique Baleriola, and Luis Gonzalez. "Dropout in youth and adult education: a multilevel analysis of students and schools in Chile." Frontiers. last updated June 1, 2023. https://www.frontiersin.org/articles/10.3389/fpsyg.2023.1163088/full.

Doyne, Shannon, and Natalie Proulx. "How can schools engage students who are at risk of dropping out?" The New York Times. Last updated: March 13, 2023. https://www.nytimes.com/2023/03/13/learning/how-can-schools-engage-students-who-are-at-risk-of-dropping-out.html.

Dufour, Raphaëlle. "Self-perception among students with learning difficulties." ÉducoFamille. last updated August 24, 2023. https://educofamille.com/perception-de-soi-eleves-en-difficulte/.

Girard, Dorian. "Behrière le décrochage scolaire, de fortes inégalités sont à l'œuvre." L'Etudiant. L'Etudiant. last updated June 2, 2023. https://www.letudiant.fr/lycee/derriere-le-decrochage-scolaire-de-fortes-inegalites-sont-a-l-oeuvre.html.

Gutierrez, Armando. "Understanding the factors contributing to teen high school dropout." LinkedIn. Last updated March 25, 2023. https://www.linkedin.com/pulse/understanding-factors-contributing-teen-high-school-gutierrez-ed-d-.

Ktourza, Sandra. "51% of parents are afraid their child will drop out of the school system." VousNousIls. last updated September 15, 2023. https://www.vousnousils.fr/2023/09/12/51-des-parents-ont-peur-de-voir-leur-enfant-decrocher-du-systeme-scolaire-677855.

Matz, Sandra C., Christina S. Bukow, Heinrich Peters, Christine Deacons, Alice Dinu, and Clemens Stachl. "Using machine learning to predict student retention from socio-demographic characteristics and app-based engagement metrics." Nature. last updated April 7, 2023. https://www.nature.com/articles/s41598-023-32484-w#Sec19.

Samy Selim, Kamal, and Sahar Saeed Rezk. "On predicting school dropouts in Egypt: a machine learning approach." SpringerLink. last updated January 12, 2023. https://link.springer.com/article/10.1007/s10639-022-11571-x.

"Why students drop out (and how to avoid it)." Embark. last updated August 8, 2023. https://www.embark.ca/learning-centre/why-students-drop-out.


See more articles by this author

Files

  • Taking risks

Thot Cursus RSS
Need a RSS reader ? : FeedBin, Feedly, NewsBlur


Don't want to see ads? Subscribe!

Superprof: the platform to find the best private tutors  in the United States.

 

Receive our File of the week by email

Stay informed about digital learning in all its forms. Great ideas and resources. Take advantage, it's free!