– High school course taking in science and performance on science and math standardized tests were significantly and positively related to an increased interest in STEM.
– More high school science courses equated to a greater likelihood of an early STEM interest, while more English decreased one’s odds.
– College aspirations were significant, and those with loftier educational goals were generally more likely to plan to major in a STEM field.
– Male high school students were significantly more likely to have an early interest in STEM relative to their female peers, as were African American high school students compared with White students.
– There was no statistically significant difference between Latinos and White students.
– Low-income students were significantly more likely to be interested in STEM majors than higher income students, respectively.
– In terms of school-level context, while teacher academic qualifications had a negative but significant relationship with an early interest in STEM, teacher experience had a small but significant positive relationship.
– Students from schools with smaller class sizes had higher odds of an early STEM interest as compared with those from schools with larger class sizes.
2013 - Predicting High School Students’ Interest in Majoring in a STEM Field: Insight into High School Students’ Postsecondary Plans
In terms of theoretical perspective, the framework uses multiple perspectives. From the rational-behavior perspective, decisions to enroll in college and the selection of majors are largely made on the basis of perceived economic returns. The sociological perspective of negative selection purports that such enrollment decisions and the selection of majors are not only governed by considering economic returns, but also cultural norms and expectations.
In the given state, all high school juniors are required to take the ACT as part of the state’s achievement test battery. The resulting dataset was a census of the graduating high school class of 2003. Immediately prior to the ACT test administration, students completed a survey called the ACT Student Interest Inventory. The analysis presented focuses on students who maintained a planned college major during high school.
A sample (n=27,935) of the cases with complete information (N=59,618) on all of the variables analyzed in the study was randomly selected. A two-level hierarchical generalized linear model with students nested within schools was initially tested. However, the null model indicated that there was insufficient variance between schools in the average likelihood of students to have an early STEM interest to justify the use of a nested model.
DV: Having an interest in STEM.
IV: Gender; Race; ACT scores; High School GPA; whether completed core curriculum; number of semesters in math, science, and English; whether or not they completed AP science; need help studying; need help in math; highest degree expected; type of college preference; family income; expected to receive financial aid; expected to work during college; number of sibling; high school program (CTE, college prep, or general); mean ACT composite for class; mean class size; attendance rate; mobility rate; teacher academic capital; average years of teaching experience.