MATH AND SCIENCE PERFORMANCE ON READING COMPREHENSION: ASYMBOLIC REGRESSION ANALYSIS

Abstract

Traditional techniques for analyzing the functional relationships between a set of predictors (x)
and a response variable (y) assure linear models. In reality, however, such relationships are
highly complex and non-linear. This study pursues a non-linear analysis of the relationship
between math and science performance (x) and reading comprehension (y) using symbolic
regression. Data from PISA (2015) were obtained and analyzed. Results revealed that while
there is no consistency among students of different countries on their dominance over a subject
area, their scores in the three areas as math, science, and reading comprehension are very close
to each other. This means that for these students, the ability to perform in math is almost the
same ability they can show in science and reading comprehension respectively and vice versa. It
was found out that the students’ performance in both math and science were highly correlated
with their reading comprehension. Moreover, the combined math and science performances of
the students had a very significant correlation with their reading comprehension. Thus, reading
comprehension is highly correlated with the performance in math and science.