Decision-making mechanisms have been extensively studied in two-alternative forced choice paradigms, where participants have to decide between two categorization options, generally located at the upper left and right corners of the computer screen. Such paradigms thus force participants to decide between two contrasted - and sometimes extreme - answers. However, these extreme response options don’t allow reporting more nuanced decisions, particularly likely for certain attitude objects or moral assertions. Furthermore, from a situated social cognition perspective, the very features of our environment, and how we interact with them, shape social cognitions, resulting in the possibility that constraining response options in such a way may have a direct influence on cognition and constrain it early on in the decision making process. In this research, we set up two experiments in the domain of morality where hand-initiated computer mouse movements were recorded, and where we presented two different response modes for reporting decisions: a dichotomous one, as in typical two-alternative forced choice paradigms; and a continuous one, where participants could also report their answers anywhere between the two extremes of the response scale. Stimuli were moral assertions adapted from Pärnamets et al. (2015) and response modes was a within-participants variable. Experiment 1 (N=104) relied on a traditional mouse tracking (MT) design (Freeman & Ambady, 2010), where each trial started with the mouse cursor at the bottom of the screen and responses were given at the top of the screen. Two-dimensional trajectories were thus recorded, as is typical in MT. Experiment 2 (N=65) addressed several limitations of experiment 1, and used a newly developed MT-like design, with a slider that had to be moved with the computer mouse along the horizontal axis of the screen only, resulting in unidimensional trajectories. Notwithstanding these differences in design and related measures, both provide access to the dynamics of the decision-making process through mouse trajectories, which could then be compared across response modes. We observed similar results in both studies: from the very beginning of the mouse trajectory, and therefore of the decision-making process, the binary mode of response leads participants to move in a more extreme way than in the continuous response mode, a non-trivial result when the explicit final response is identical across conditions. We discuss implications for morality, and present preliminary findings from computational models of paradigm dependent decision-making dynamics.