Our paper on predicting voting decisions using network analysis was published in Scientific Reports (PDF) and our tutorial on analyzing and simulating attitude networks was published in Social Psychological and Personality Science (PDF).

In the first paper, we show that whether attitudes toward presidential candidates predict voting decisions depends on the connectivity of the attitude network. If the attitude network is strongly connected, attitudes almost perfectly predict the voting decision. Less connected attitude networks are less predictive of voting decisions. Additionally, we show that the most central attitude elements have the highest predictive value.

In the second paper, we provide a state-of-the-art tutorial on how to estimate (cross-sectional) attitude networks and how to compute common network descriptives on estimated attitude networks. We also show how one can simulate from an estimated attitude network to derive formalized hypotheses regarding attitude dynamics and attitude change.