This chapter encourages to social psychologists to solidify and advance their understanding of social influence and collective behaviour by creating and testing computational models—models that can be formally stated with enough specificity to allow them to be run as computer simulations. Almost a century ago, Floyd Allport suggested that there are two kinds of data in social psychology: the behaviour of an individual in direct response to social stimulus; and behaviour which is the response to a non-social stimulus. The chapter focuses efforts to make progress on that problem—specifically, the nature of the relationship between the opinions and behaviours of individuals and those of others around them. Wilder has demonstrated that marginal influence drops off more rapidly as a function of how similar sources are to each other. Davis proposed a more general modelling framework for social-psychological research on group decision processes, one that has been quite productive for many decades.