Thinking is the essence of human existence, unquestionably the highest product of human evolution. Yet it is not possible to convey coherently to students the sum and attainment of all humanity's knowledge. Students would be incapable of absorbing such an enormous quantity of knowledge (corresponding to current developments in individual academic disciplines), as this is often highly abstract. Therefore educational content must be selected, to the extent that it presents a representative selection of knowledge. One of the most effective selection methods could be the use of conceptual mapping in the educational process. In this study the authors focus on the issue of structured teaching material and its relational association with creating students' mental representation, and have several objectives. The first is to analyze constructivist-oriented teaching processes; the authors also give attention to basic theses by which it is possible to delineate the main pillars of constructivist-oriented teaching. Attention is also given to learning strategies. No less important is the analysis of non-linear structured teaching material and its specific types, with an emphasis on conceptual mapping and its role in the teaching process -- in the empirical part the effect of non-linear material in recording topical unit meaningfulness in students' mental representation. In the research undertaken, the methods of conceptual mapping experimenting and testing were used. The consistency of the conceptual maps was evaluated through the relational method using ordinal variables: meaningfulness of conceptual maps. The authors assessed the quality of the conceptual maps with the aid of IRT theory, specifically the latent correlations model. The research sample comprised 96 respondents. Analysis of results showed that the use of non-linear structured teaching material improved students' results in the experimental groups in the researched operationalization parameter.