Using survival analysis to investigate how students’ ideas about structure and property relationships evolve during the first two years of college chemistry courses



Sonia M. Underwood, Melanie M. Cooper, David Reyes-Gastelum
Longitudinal studies can provide significant insight into how students develop competence in a topic or subject area over time. However, many aspects of longitudinal studies can become problematic such as retention of students in the study, financial abilities to continue the project, and difficulty with the data analysis process. To address the concern of data analysis, we chose to use a discrete-time survival analysis to investigate how students ideas about the connection between the molecular-level structure of a substance and its macroscopic properties develop over the first two years of introductory college chemistry. To evaluate students’ ideas about this connection, we have administered the Implicit Information from Lewis Structures Instrument (IILSI) to capture the types of information students believe can be predicted using a Lewis structure. This instrument was administered five times over a two-year period (i.e. two semesters of general chemistry and two semesters of organic chemistry. In addition, it was administered to three different cohorts of students to see if the results could be replicated with two subsequent years of students. A description of discrete-time survival analysis along with the results from this study will be presented.



File Attachment: