Measuring Effects on Psychological Networks

What is it about?

Network models are useful tools in psychology because they help us understand how different factors interact with each other. But when we try to compare these networks across different groups, it can get complicated, especially with large amounts of data. This study offers a solution by connecting network models with something called item response theory (IRT), a method often used in testing and measurement. We show that a specific IRT model, called the two-parameter logistic model, can help us find out how causes affect both the whole network and the strength of the relationships within it. Using computer simulations, we show that this method works well. To show how this can be used in real life, we first apply our method to vocabulary test data from a literacy program. Then, we apply it to data from 72 different experiments in areas like education, economics, and health. We find that interventions (like new teaching methods or programs) often change how strongly things are connected in the network, but don’t always change the overall state of the network. Our study gives researchers a practical way to better understand how interventions work, especially in social and behavioral sciences.

Featured Image

Photo by JJ Ying on Unsplash

Photo by JJ Ying on Unsplash