The inequity aversion model is fundamental to explaining the motivation of voluntary contribution behavior. Yet, the empirical studies limit the research on the inequity-aversion to small groups due to the difficulty of accounting for each individual’s payoff in the decision-making process for large group sizes. To overcome the technical barrier of estimating expected payoffs with large group sizes and various marginal returns, we adopt the Quantal Response Equilibrium estimation method that allows for an individual’s contribution to be impacted by their social preferences, i.e., inequality aversion, altruism, and bounded rationality. Our theoretical model identifies the equilibrium strategy of contribution. Using Bayesian inference via Markov Chain Monte Carlo, the experimental results show people are significantly worse off when they contribute more or less than others, meaning social equality promotes public good provisions. We find that marginal per capita return and group size play important roles in people’s contribution decisions, and the effect of marginal per capita return on contributions is stronger than that of group size.