While many solutions to the apparent civic online reasoningdeficit have been put forth, few consider how reasoning is of-ten moderated by the dynamic relationship between the user’svalues and the values latent in the online content they are con-suming. The current experiment leverages Moral FoundationsTheory and Distributed Dictionary Representations to developa method for measuring the alignment between an individual’svalues and the values latent in text content. This new measureof alignment was predictive of bias in an argument evaluationtask, such that higher alignment was associated with higherratings of argument strength. Finally, we discuss how theseresults support the development of adaptive interventions thatcould provide real-time feedback when an individual may bemost susceptible to bias.