The correlation regression equations between metacognitive skills and concept gaining of science and biology of Indonesian students

Abstract

In the field of education metacognitive skills can be used as predictors of students’ concept gaining. There have been many researches on the correlation between metacognitive skills and concept gaining both at the elementary school level and high school levels. Several further researches were also conducted to uncover the factors that influenced the regression lines of the correlation. However, to date the actual factors affecting the coefficient value of the slope and intercept of the regression equations on the correlation between metacognitive skills and concept gaining have not been revealed. This research aimed at investigating the factors affecting the slope coefficient and intercept of the regression equation of the correlation between metacognitive skills and concept gaining of students in different educational levels implementing of one learning model. The design of this research was a correlational research in order to investigate the correlation between metacognitive skills as the predictor and the concept gaining as the criterion. Among the 27 regression equations related to the correlation between metacognitive skills and concept gaining of the students undergoing various learning models, that have been analyzed, 77.50% of them are parallel, and only 22.50% of them are intersected. Although there were some intersecting lines, the percentage of parallel lines was bigger. Several factors (education level, class level, academic ability, and learning models) which had been predicted to affect the slope coefficient as well as the intercept were finally proved to be non-affecting factors, because both parallel regression lines and intersecting regression lines were still found in the various conditions compared. This research revealed two factors affecting the slope coefficient and the intercept. These factors are the presence of shared variables that affect the value of X and Y and the measurement error. Further researches related to these factors are necessary to be carried out to uncover more accurate results.