The main objective of this study was to evaluate the effect of teaching experiments within a dialectic-constructivist framework based on the following considerations: (a) Cognitive conflicts used in the teaching experiments must be based on problem-solving strategies that students find relatively convincing: (b) after having generated a cognitive conflict, it is essential that the students be provided with an experience that could facilitate the resolution of the conflict; and (c) the teaching strategy developed is used by an interactive constructivist approach within an intact classroom. The study was based on two sections of freshman students who had registered for Chemistry I at the Universidad de Oriente, Venezuela. One of the sections was randomly designated as the control group and the other as the experimental group. To introduce cognitive conflict, the experimental group was exposed to two teaching experiments dealing with stoichiometry problems based on the concept of limiting reagent. Students in the control group were exposed to the same problems—however, without the cognitive conflict teaching experiments format. To evaluate the effect of the teaching experiments, both groups were evaluated on five different problems at different intervals during the semester, referred to as posttests. All posttests formed part of the regular evaluation of the students. Results obtained show the advantage of the experimental group on four of the posttests. It is concluded that the experimental treatment was effective in improving performance on the immediate posttests. It was observed that some students protect their core belief [see Lakatos, I. (1970). Falsification and the methodology of scientific research programmes. In I. Lakatos & A. Musgrave (Eds.), Criticism and the growth of knowledge (pp. 91–196). Cambridge: Cambridge University Press] in stoichiometry (establishing equivalent relations between different elements or compounds) by ignoring the conflicting data, just as conflicting (anomalous) data do not necessarily convince a scientist to abandon a particular theory.