Evidence of tutoring, mentoring and support programs for the improvement of reading and mathematics competences.

How to cite

Calsamiglia, Caterina; Giorgi, Giacomo De; Viviano, Davide; Ruiz-Valenzuela, Jenifer; Garcia-Brazales, Javier. 2022. "Evidence of tutoring, mentoring and support programs for the improvement of reading and mathematics competences.." AEA RCT Registry. AEARCTR-0010448. https://doi.org/10.1257/rct.10448-2.0

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Caterina Calsamiglia* , Giacomo De Giorgi , Davide Viviano , Jenifer Ruiz-Valenzuela , Javier Garcia-Brazales

* Primary investigator

Registry ID
AEARCTR-0010448
DOI
https://doi.org/10.1257/rct.10448-2.0
First registered
Last updated
Status
In development
Trial dates
2022-11-23 – 2024-03-31
Countries
Private
JEL codes
I24, J18
Registry record
http://www.socialscienceregistry.org/trials/10448

Abstract

Reading and math skills have a direct impact on people's lives. Lacking these skills constitutes an important barrier against effective communication, and therefore against the development of the individual within society. An early deficiency in reading and mathematics levels is correlated with lower academic performance, a higher level of dropout, and worse labor market outcomes.

Resolving these deficiencies at early ages has two main potential benefits. First, it unravels the development of the person, making it easier for him/her to progress in their learning (dynamic gains and complementarities of future investments). Second, it improves the ability to work in the classroom on the part of the whole group (spillovers). The sooner such gaps and inequities are identified and addressed, the better individual and classroom educational performance will be.

The available evidence, generated from intervention projects developed in countries other than Spain (our context of interest), points to tutoring and mentoring programs as one of the most effective means to increase the academic performance of students, especially those who start at a greater disadvantage. In addition, the existing evidence points to this type of intervention as triggering other positive effects on learning (personal and transversal competences, school adherence, reduction in absenteeism, etc.).

The objective of this study is to provide evidence about the effectiveness of two interventions providing educational support, one for reading and one for mathematics, in low-socioeconomic schools located in Spain. The reading program consist of one-on-one mentoring sessions for grade 4 students aimed at fostering the pleasure for autonomous reading, The mathematics program consists of four-group tutorial sessions for grade 6 students. Within the classes of the grade for which the school is treated, only a subset of students will actually participate in our tutoring/mentoring programs.

Assigning half of the pool of participating schools into the reading program and the other half into the mathematics program, we will compare the outcomes of the students from the grade for which the school has been treated with the outcomes of the students in the same grade from schools that got assigned to the other program. Within a given treated class, students who did not actually participate in the mentoring/tutoring programs might still have their outcomes impacted through in-class spillovers.

Naturally, we will explore the impact of our intervention on reading and mathematics cognitive abilities. We will pay particular attention to how these programs affect the classroom depending on the position in the network of friends where the treated kids belong. We also explore other dimensions such as improvements in socio-emotional skills, classroom climate and behavior, and changes in social relations and friendships.

Experimental design

We will evaluate the impact of two programs targeted at primary education students in disadvantaged areas that take place outside school hours. One is a reading mentoring program for Grade 4 students consisting of one-on-one sessions with a volunteer. The other one is a mathematics tutorial program delivered in groups of four among Grade 6 students.

We have committed to implement one of the two programs in each and every of the schools that have agreed to participate in our study. For this, we randomly assign half of the schools to the reading treatment and the other half to the mathematics treatment. Within each treated class, only a subset of students will participate in the program due to budgetary reasons. We will identify the effects of a given program by comparing the outcomes of the students in the relevant grade in schools that were assigned to treatment for that program (some of these students received the treatment directly and others through classroom spillovers) and the outcomes of the students in that grade from students of schools that got assigned to the other treatment. These control students did not receive the treatment in any form (we assume that cross-grade spillovers are non-existent).

Design details

Randomization method
The randomization algorithm was written by the PIs and then executed (including the selection of the seed for random number generation) under the supervision of our implementation partner, Fundació Bofill.
Randomization unit
Randomization takes place at two levels. In the first one, the pool of participating schools gets split into two: those that will receive the reading treatment and those that will receive the mathematics treatment. In the second level, a subset of the students in each class gets assigned to participate in the program. For the selection of these students, their headteacher provided a ranking based on how he/she perceives the needs of each student for participating in our program to be. We follow this ranking until completing the pre-determined number of participants per class/program and only intervene by randomizing which student to receive treatment in the case of oversubscription within a class of students with the same level of perceived needs.
Clusters
A total of 134 schools initially agreed verbally to participate in our study. A subset of them might nevertheless decide not to proceed with the signing of a formal agreement of participation and will then be excluded from our investigation. The allocation of schools to treatment arms was performed before knowing which schools will decide to discontinue their participation.
Observations
Within each school, all students in grades 4 and 6 belong to our target population (as explained, only a subset of them actually receives our treatment). If all the 134 schools participated and all the students in those grades did so as well, approximately 10,000 students would be expected to participate in our study. This number is, however, prone to be significantly lower due to multiple reasons, including: 1) student absenteeism on the days of information collection; 2) students' inability to speak the language; 3) lack of parental consent; 4) schools eventually deciding not to participate as a whole.
Arms
50% of our schools will be assigned to the reading treatment and 50% to the mathematics one. This amounts to 67 schools in each treatment.
Primary outcomes
Academic scores (mathematics and reading); non-cognitive measures (academic interest and aspirations, locus-of-control, self-confidence), class environment and friendship networks.

Cited by

No citation data available for this registration yet.

Keywords