Award Date

May 2024

Degree Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Teaching and Learning

First Committee Member

Hasan Deniz

Second Committee Member

Merryn Cole

Third Committee Member

Tina Vo

Fourth Committee Member

Mingon Kang

Number of Pages

117

Abstract

The dearth of STEM students in the United States is a growing concern for policymakers and educators alike. With the increasing reliance on technology in the global economy, a STEM trained workforce is essential for the United States to remain competitive. However, the number of students majoring in STEM disciplines and pursuing STEM careers is not keeping pace with the demand for these skilled workers. As a result, understanding the characteristics that contribute to students' confidence in science and their desire to pursue professions in science remains a national priority. This research investigated the factors influencing the choice of STEM careers among high school graduates. To achieve this, the study analyzed data from 520 high school graduates, using machine learning models and chi-square analysis to predict their propensity for choosing STEM careers. This study evaluated the performance of four machine learning models—Logistic Regression, K-Nearest Neighbors (KNN), Decision Tree, and Neural Network—across various metrics, including accuracy, precision, recall, and F1 Score, to determine their effectiveness in classification tasks. The Logistic Regression model has a performance with an accuracy of 83.65%, precision of 88.83%, and recall of 84.12%. This indicates a slight preference for precision over recall. On the other hand, the K-Nearest Neighbors (KNN) model shows better accuracy (87.5%) and recall (93.6%), but with a slightly lower precision (86.7%). This suggests that the KNN model is effective in identifying relevant instances but with some compromises in precision. This study also explored the impact of socioeconomic status (SES), ethnicity, and access to Advanced Placement (AP) courses on the career choices of students in STEM fields. It found that gender did not significantly affect these decisions, but disparities in SES, ethnicity, and educational opportunities played a critical role. The study recommended that educational stakeholders work together to address these disparities by providing supportive measures and equitable resource allocation to promote a more inclusive and diverse STEM workforce.

Keywords

Machine Learning; Science Education; STEM

Disciplines

Curriculum and Instruction | Curriculum and Social Inquiry | Education | Educational Methods | Science and Mathematics Education

Degree Grantor

University of Nevada, Las Vegas

Language

English

Rights

IN COPYRIGHT. For more information about this rights statement, please visit http://rightsstatements.org/vocab/InC/1.0/

Download

Available for download on Thursday, May 15, 2031


Share

COinS