Statistical Modeling of the Factors Affecting Mathematics Performance in Secondary Schools: A Case of Kigoma Region
Loading...
Date
2020-10-04
Journal Title
Journal ISSN
Volume Title
Publisher
MUST Journal of Research and Development (MJRD)
Abstract
This work was based on analysis of factors affecting mathematics performance for the purpose of
developing multiple linear regression models by using Ordinary Least Squares Method. The data was
collected in Kigoma region from five selected secondary schools. This study analysed students’
observed performance data of 92 and 89 selected samples for both basic mathematics and advanced
mathematics respectively. Principal Component Analysis method was used to reduce 30 Likert scaled
observed variables and filter out latent factors which affect mathematics performance. Latent
independent factors namely; students’ self-efficacy and self concept towards mathematics, students’
engagement time in mathematics, students` motivation and evaluation strategies, mathematics teachers
and resource availability, students’ mathematical intelligence and access to learning materials, and
the school mathematics learning environment were found to be significant factors that influence
performance in either Basic or Advanced mathematics. Sensitivity analysis was carried out by
construction of partial regression plots and standardized regression coefficient rank approaches. For
both basic mathematics and advanced mathematics, students’ performance was predicted using
regression models that incorporated independent factors which explained 88.6% and 66.4% of
variance in mathematics performance for each model respectively. All significant factors were tested
using t-test and found that each factor have positive correlation with mathematics performance.
Finally using 95% confidence interval, the simulation of predicted performance was done using simple
scatter plots for each model. The scatter plots verified that the predicted mathematics performance
reflect real life situation at 0.05 level of significance.
Description
This Journal Article was Published Mbeya University of Science and Technology, University of Dar es Salaam