A research was conducted on an educational proposal of learning to teach science through self-regulation in which chemistry teachers-to-be, teamed up in groups of two, think about assessment objects and criteria (supposedly agreed upon) during their Teacher Training Course carried out at a secondary school in Neuquén city - Argentina. It focused on linking the motivational-emotional process of each of the four selected Cases with the promotion of changes of teaching-learning goals / models, thus analysing their discourse obtained through Questionnaires, Interviews and their Reports. The complex system of holistic categories considered their scientific, didactic and emotional knowledge. From the paradigm of the professor-to-be professional development and personal growth, we approached emotion from Perspectives: Post-structuralist, Neurobiological and Psychological (Cognitive) which links secondary emotion-motivation and cognition. The teachers-to-be pursued more than one goal such as that of feeling secure. Group I, more successful (experts), always pursued goals associated with learning by showing some theoretical fundamentals, intuition close to that of the tutor-researcher's and favourable and very favourable emotions for learning. The latter boosted self-regulation in the action of some not very favourable difficulty-emotion such as the change in the organization of school scientific Knowledge. Reflections with no evidence of emotion showed the concealment of others. Group II, less successful (an expert and a novice learner), pursued goals associated with the self, which generate very little favourable or unfavourable emotions. The latter only enabled them to self-regulate as possibility and at the end of the process (Reports) of any difficulty-emotion such as the so self-focused Communication. All of them improved mutual assessment, help and self-reflection, and integrated reason/emotion. Emotional changes did not always entail changes of goal/model depending on the type of cause awarded to the difficulty-emotion and of strategy used for sorting it out. Four styles of emotional work were defined: altruistic, attractive, overwhelmed and phlegmatic, as well as hypotheses for working with each one. Overcoming the burden of their experience represented, in the most successful ones, one of the various causes of satisfaction, limited, by the way, in the least successful ones. New studies are proposed by improving the sample and for detecting other emotional work styles, whereas in such proposal, the aim is to insist in enhancing the attribution interval that professors-to-be grant to their results so they will start changing their success and failure perceptions.