We investigate how much of a person’s deep moral preferences can be retrieved from observing their choices, for instance via experiments, and in particular how one should interpret behaviors that appear deontologically rather than consequentially motivated. Comparing the performance of the direct elicitation (DE) and multiple-price list or Becker-DeGroot-Marschak (BDM) mechanisms, we characterize in each case how (social or self) image motives inflate the extent to which agents behave prosocially -e.g., refuse “bribes” for causing harm. More surprisingly, the signaling bias is shown to depend on the elicitation method, both per se and interacted with the level of visibility: it is greater under DE for low enough reputation concerns, and greater under BDM when they become high enough. We also provide conditions ensuring a single crossing. We further show that, even when all agents are consequentialists, certain Kantian behaviors and postures easily emerge under BDM (but not DE) when reputation becomes important enough, with both high and low-morality agents turning down all prices within the offered range.