Open Access Te Herenga Waka-Victoria University of Wellington
thesis_access.pdf (10.33 MB)

The Analysis of Samoan Schools Dropout Rates

Download (10.33 MB)
posted on 2021-11-12, 22:11 authored by Fitu, Lealaolesau

This thesis investigates the dropout problem in Samoan schools particularly at primary and secondary levels from 1995 to 2007. It aims at identifying and comparing the dropout rate by region (or geographical locations), school level (primary, primary/secondary and secondary) and school status (Government, Mission and Private). Moreover, it also investigates whether the student-teacher ratio, school size, the gender and ethnicity of the teacher, the qualification of the teacher, the school building and school facility variables are associated with the dropout of students. The investigation is carried out through analysis of census data gathered annually by the Samoa Ministry of Education, Sports and Culture (MESC) through census forms from all the primary and secondary schools in the country. Given our response variable is a dichotomous one, the logit regression models to model the effect of both the categorical and continuous explanatory variables on the dropout was adopted. Moreover, since the dropout rates across different year levels (eg, Year 2, Year 3, and Year 13) are correlated within each school we then use the Generalised Estimating Equations (GEE) approach. The results show that higher dropout rates are found in the rural areas (Rest of Upolu combined with Savaii region) as compared to urban area (Apia urban region). In addition, students are more likely to leave secondary schools earlier than primary or primary/secondary schools. Apparently, the majority of these dropouts are those who have attended Government schools. Surprisingly, students are less likely to get affected with the dropout in larger schools and those schools with higher student-teacher ratios. The gender of the school teacher has nothing to do with the dropout however; having more Samoan and highly qualified teachers in a school will significantly reduce the dropout rate. Nonetheless, a couple of school building variables are significantly associated with the dropout in the positive direction, while another couple inversely relate with the dropout. Schools with more trucks as means of transportation for students, and more radios for school programs and students’ activities are less likely to get affected with the dropouts.


Copyright Date


Date of Award



Te Herenga Waka—Victoria University of Wellington

Rights License

Author Retains Copyright

Degree Discipline

Statistics and Operations Research

Degree Grantor

Te Herenga Waka—Victoria University of Wellington

Degree Level


Degree Name

Master of Science

Victoria University of Wellington Item Type

Awarded Research Masters Thesis



Victoria University of Wellington School

School of Mathematics, Statistics and Operations Research


Liu, Ivy