In this study the answer is parents, but I do think this study needs replication in different regions to see how it differs based on different cultures. I can also imagine other elements playing a role that weren’t included in this study.
From the press release:
A new study shows that the elementary school a child attends has almost no influence on their desire to progress to higher education – as factors including parental aspirations, academic support from their mother and having a desk to work on are much more important.
Published in the journal Educational Studies, the findings of the research looking at 1,000 pupils showed that school and class size, the grade point average of the school and property prices, had little influence on the desire to continue to higher education.
The research was carried out by Josip Šabi? and Boris Joki? at the Centre for Educational Research and Development of the Institute for Social Research in Zagreb, Croatia, and was supported by the Croatian Science Foundation. The authors wanted to discover the main factors affecting pupils’ intention to continue to higher education as they reach the end of elementary school.
In Croatia, children attend elementary schools up to age 14-15, at which point they move on to a secondary school. Here, they can either study for a four-year diploma, after which they have the option of applying for university, or a three-year diploma, which prepares pupils for work but does not permit them to apply for university.
To find out children’s aspirations, they asked just over 1,000 pupils at 23 elementary schools in Zagreb to complete three separate questionnaires during their last two years at elementary school. These questionnaires asked them whether they would like to continue to higher education, as well as about their parents and home life. There were questions about their parents’ aspirations for them, the level of academic support they received from each of their parents, whether they had their own room, computer and desk, and whether they enjoyed school.
The researchers also obtained information on the pupils’ academic grades, as well as on the size of each school and its classes, the grade point average for each school, and property prices in the area around each school as a measure of socioeconomic status. Finally, they performed statistical analyses on these responses to determine which factors were most closely related with a desire to continue on to higher education.
This revealed that none of the school-level factors, including school and class size, grade point average of the school and property prices, had any influence on the desire to continue to higher education. In contrast, several factors related to parents and home life, such as parental educational aspirations, maternal academic support and having a desk to work on, did have an influence. As did gender, with girls more likely than boys to want to continue to higher education. And while school-level factors didn’t have any influence, performance at school did: high academic grades were the single strongest predictor of a pupil’s desire to continue to higher education, while enjoying school was also an important factor.
“The major finding arising from the present study is that none of the school level variables used in our analysis contributes to the explanation of pupils’ aspirations for higher education,” said Šabi?. “In other words, pupils who have similar individual characteristics but attend different schools will likely hold similar aspirations for higher education.
“Another important finding is that parents can influence their child’s aspirations by expressing their expectations regarding the child’s educational path and by providing the basic conditions for completing homework and learning (i.e. a desk to work on).”
Abstract of the study:
This paper discusses the contributions of variables commonly used within the status attainment and blocked opportunities models in addition to school context variables in predicting pupils’ aspirations for higher education prior to the transition to differentiated upper secondary education in Croatia (14–15 years). The findings, arising from a longitudinal design, indicate that the school context does not contribute to the explanation of pupils’ aspirations for higher education. In contrast, gender, parental educational aspirations for their child, maternal academic support, the pupil’s possession of a work desk, pupil grades and school satisfaction predict aspirations. Arguably, pupils who attend different schools but have similar individual characteristics will likely have similar aspirations for higher education.