This study reports quantitative evaluation of changes in students' study approaches during the first semester of organic chemistry. Students were surveyed by a modified M-ASSIST (OCH-adjusted M-ASSIST) instrument adjusted to the specific student population. The validity of the data-model fit was assessed by confirmatory factor analysis. Students were surveyed twice per semester, and 93 individually paired responses afforded pairs of deep and surface learning scores (number on 1-5 scale) for each respondent on each sampling date. The statistical significance of the changes in the two mean learning scores between the two sampling dates was assessed separately for the four achievement level (final grade) groups. Statistically significant differences were observed for changes in deep learning scores for groups A and B, suggesting an increased adoption of the deep learning approaches, and in surface learning scores for the group C, indicating an increased reliance on surface learning approaches. The results revealed a "characteristic pattern of evolution in the mean learning scores" for the achievement level groups A/B versus C versus D/F. Students in higher achieving groups (A and B) increased their adoption of deep learning approaches without a change in their surface learning scores, whereas scores for students in the lowest achieving group (D/F) did not change. Notably, scores for students in the achievement level group (C) indicated an increased adoption of surface learning approaches. The study provides a valuable insight into how students of each achievement level modify their learning approaches in response to the demands of the organic chemistry course.