FESE HandBook v03c 15112023 MEV- COMPLETO - Flipbook - Página 399
ROBBI E DEAN + AGNES TOLESCU / WHOLE CHILD DEVELOPMENT IN THE TEACH FOR ALL NETWORK
The study is a quasi-experimental impact evaluation that uses a
mixed-methods approach to analzye student and teacher level data from
both comparison and TFN classrooms, and triangulates these ondings
with qualitative data from key informant interviews of school community members. The baseline and follow-up data collection for the impact
study involved three types of respondents to self-reported surveys (headteachers, teachers, and students) designed to measure whole child development outcomes of students and factors that innuence their development. Students were also administered grade-specioc mathematics and
literacy assessments based on Ogun State curricular standards developed
by Ogun State district education ofocers. Social and emotional learning
skills were measured through a triangulation methodology, relying on
both student self-reported questionnaires based on internationally validated student social and emotional learning scales and from perceptions
of teachers, parents, and school leaders in qualitative interviews. The data
collection was conducted by Hanovia Limited in two waves; baseline
data collection occurred in October-November 2021 and follow up data
collection occurred in May-June 2022. The qualitative data collection
was performed in May-June 2022, and consisted of in-depth interviews
and focus groups. Additional data will be collected at the end of the 2023
school year to understand more about the two-year effect of TFN fellows
on whole child development, classroom environment, relationships, and
school-level transformations.
The total sample from the 2021 and 2022 data collection periods consisted of students, teachers and headteachers who attend schools in Ogun
state, separated between 40 public primary schools where TFN fellows
are placed, and 40 similar comparison schools that were selected using a
8matched pairs9 approach to ensure that the treatment and comparison
schools serve a similar student population. Data was collected from a total sample of 80 headteachers, 180 teachers and 5900 students in grades
3-6 in Ogun State. Grades 3-6 were selected as the focus of the study given
the overall availability of TFN teachers teaching these grade levels and
the applicability of student assessment to these grade levels. For the qualitative study, the sample consisted of eight focal schools, where 19 TFN
fellows, 8 headteachers, 16 non-TFN fellows, and 67 parents of students
taught by TFN fellows participated in interviews and focus groups.
Quantitative student data was analyzed using a difference in difference regression modeling approach, with two time periods and a single
treatment assignment of whether students were taught by a TFN teacher, as well as a rich set of control variables measured at baseline. The
treatment effect was estimated using an ordinary least squares model,
using sandwich-estimated standard errors to account for the clustering
399
