A step-selection function (SSF) is a powerful method which quantifies the environmental influences of animal movement and is important in identifying conservation strategies associated with landscape connectivity. A SSF is calculated using conditional logistic regression to differentiate between environmental factors associated with a step actually taken by an individual compared to a set of potential random steps. This method is increasingly being applied to biogeographical questions, but there has been little research on how parameters specified in the model building phase affect the results. This research quantitatively assesses how user decisions influence the coefficients of the subsequent regression models. Systematically altering the method of generating available steps, the number of steps used, the number of individuals in the model, the method of conditional logistic regression (population v individual), we found that user decisions strongly influence the results of step-selection functions and any subsequent inferences about animal movement and environmental interactions. The largest significant differences occurred between conducting a population level regression model and individual level model, highlighting the importance idiosyncratic preferences of the individuals used in the model and the importance of the user to acknowledge such. Differences were found between categories for every variable used in analysis and will inform SSF practitioner’s with further information with which to develop SSF research and reduce uncertainty when discussing results.