Abstract:
The Government of the Republic of Kenya has continued to increase expenditure to education to enable
acquisition of learning resources. However, low learning outputs have continued to be realized every time
secondary education summative evaluation results are released. The purpose of this study was to establish the
influence of per student expenditure on learning output in science and mathematics in public secondary schools
in Siaya County, Kenya. The study was underpinned on the Education Production Functions Model by Hanushek
(2008). The study employed convergent parallel mixed methods design. The target population for this study was
6175 comprising 247 principals, 988 teachers and 4940 students from the 247 public secondary schools in the
study locale. A stratified random sampling technique was employed to select 376 respondents was used from 15
sampled schools. The average termly class mean scores were used to measure the students’ learning output in the
schools. Questionnaires, interview guides and document analysis guide used to collect data on per student
expenditure and learning output. Test re-test method was used to determine reliability of the instruments. The
quantitative data was analyzed using descriptive and inferential statistics and presented in tables and graphical
format, while the qualitative data was analyzed thematically and presented in quoted texts. The study found a
positive linear relationship between per student expenditure and learning output. The coefficient of
determination R2 was established at 61.9%. In conclusion, schools which had appropriate resource inputs to meet
quality learning requirements also had better learning output. Though nearly all the study schools fall short of the
policy standard requirements with the predictor variable, the more the resources, the better the learning output in
the schools. The study recommended that the national government should increase grant to cater for more school
expenditure
