An Artificial Neural Network Model For Predicting Attainment Of The 50:50 Gender Ratio In Stem Courses In Kenya

Abstract

In spite of the existing educational policies on gender and several other interventions that are aimed at empowering the girl child, education is not globally available and gender inequality is still a major problem world wide. Many nations are now concerned that fewer girls are going to school in comparison to their male counterparts, and also that males have higher participation and learning achievements than girls ,more particularly in Science, Technology, Engineering and Mathematics (STEM) subjects and courses. STEM education is one of the pillars behind Kenya’s Vision 2030, which aims to turn the country into a newly industrializing, middle-income country providing a high quality life to all its citizens by the year 2030, in a clean and secure environment. STEM education is expected to provide learners with the knowledge, skills, attitudes and behavior required for inclusive and sustainable societies. Graduation trends from the Commission for University Education (CUE) show that more than 30 % of graduating students each year are awarded commerce degrees or one of its other hybrids in business studies, 20 % graduate in education arts and another 20 % in other non-STEM courses. In a study conducted by Dr. Eusebius Juma Mukhwana, (Mukhwana et al., 2016) a former deputy commission secretary in charge of planning and research development at CUE, 74 % of all university students are enrolled in business, education arts and humanities. This leaves only 26% of the students in STEM.To make a bad situation worse, gender disparity within STEM fields is in favor of males. Female students represent only 35% of all the students enrolled in STEM- related fields of study at higher learning levels according to a study conducted by UNESCO through the ‘STEM and Gender advancement’ project in 2015. This disparity in gender is startling, moreso since careers in the STEM fields are now being commonly cited as jobs of the future that are being used, and shall continue to be used to drive innovation, inclusive growth and sustainable development. The female gender is held back by societal norms, biases and prejudice, and expectations that influence the quality of education they receive and even the subjects they choose to study at higher learning levels. Following the above findings, the Kenyan government and stakeholders in the education sector have put measures in place in a bid to bridge this gap in gender. The main aim of this study therefore was to develop a model that would predict when the ratio of males to females in STEM will be 50:50 and further determine what measures can be put in place by government or society, to promote the interest and engagement of girls in STEM. An Artificial Neural Network (ANN) was applied as the predictive data mining method to come up with the model. Exploratory data analysis was performed on the data and a regression model was built inorder to achieve the main objective of the study. The study utilized the data in the repositories of the Kenya Universities and Colleges Central Placement Services (KUCCPS) for the years 2014, 2015, 2016, 2017 and 2018. The method of data collection was ‘Use of existing data as a data collection method for machine learning’ (Yuji et al., 2019). After the model was built, it was evaluated to determine its accuracy.