Sociality is associated with a variety of costs and benefits. One potential benefit is the higher propensity of groups to solve novel problems relative to individuals alone. Several hypotheses have been put forward to explain why in some species, including humans, groups have been found to be more apt at solving novel problems than single individuals. Increased problem-solving efficiency of groups has been suggested to arise through shared antipredator vigilance and/or a pool of competence effect, whereby larger groups containing a more diverse range of individuals are more likely to contain individuals with the skills necessary to solve the particular problem at hand. Interference between group members might cause groups to have lower problem-solving efficiencies than individuals alone, however. Using a simulation approach, we modeled the shape of the relationship between group-level problem-solving probability and group size across a range of facilitating and inhibition scenarios to determine whether mechanisms underpinning group-size effects on problem solving could be disentangled. Our models indicate that measuring group-level performance as a function of group size allows facilitation mechanisms to be differentiated from inhibition mechanisms, but measuring individual performance as a function of group size does not. We show that in contrast, antipredator vigilance and the pool of competence effects cannot be distinguished by quantifying how group-level performance changes as a function of group size, nor can these mechanisms be distinguished by measuring how individual performance changes as a function of group size. Our findings have important implications for future experimental work on innovative problem solving in social groups.