In the present study we used the technique of word association tests to assess students’ cognitive structures during the learning period. In particular, we tried to investigate what students living near a protected area in Greece (Dadia forest) knew about the phenomenon of decomposition. Decomposition was chosen as a stimulus word because it represents a complex issue and was therefore suitable given the primary methodological objectives of this article. Specifically, we tried to develop a complete theoretical scheme for grasping aspects of complexity concerning the learning of biological concepts. Perhaps most importantly, we made an effort to introduce network analysis within the field of science education and evaluate its usefulness in assessing students’ knowledge structures. Network analysis was used to manage the data from the word association test and proved to be quite efficient. It became clear that such analysis may help researchers to relate cognitive structures with underlying patterns, including misconceptions. In our case study, such misconceptions stem from students’ knowledge gaps, mainly concerning holistic aspects of understanding decomposition.