Modelling is of substantial use in policy analysis. As policy analysis tries to facilitate decision making of complex multi actor problems under uncertain circumstances, it needs to rationalize and simplify reality. However, modelling is a difficult task, as the model should contain all (and preferable only) the relevant aspects of the real problem field. This requires lots of experience. During our modelling courses and the collaborative modelling workshops we perform, we noticed that students faced difficulties in modelling and run into a distinct set of pitfalls when the try to model. We defined four common and distinctive pitfalls: abstraction pitfall, labelling pitfall, conceptual pitfall and analysis communication pitfall. In this paper we give clear descriptions of each, so they are more easily recognizable. It is noteworthy that those pitfalls restrain junior analysts from delivering a thorough and qualitative analysis given that they deteriorate the analytic content-quality of the study where modelling is the core analytic method.
Given our experience in teaching modelling to graduate students, we also practice and test different types of solutions with which we guide students to overcome those pitfalls by self-evaluation of their own model and analysis. There are five recommended alternatives included in this paper for whose we provide concrete descriptions on how they are applied in practice. They are the following: (1) reiteration, (2) renaming or remaking, (3) keeping the big picture in mind and, (4) return to the roots and (5) Check and double-check.
Last but not least we aim at providing insights to teachers of modelling techniques to recognize those pitfalls (and struggles) of students so as to improve their education methods and approaches when it comes to modelling. Our systematic way of describing the pitfalls including the potential solutions will make it easier for teachers and supervisors to guide junior modellers in making better models.