prediction
Definition
Prediction refers to the use of a summary variable derived from a psychometric network model to estimate an individual's standing on a psychological outcome outside that network. The core problem is that network models do not reduce a construct to a single variable the way factor models do, which forces researchers to choose among competing network-derived metrics when modeling associations between constructs. Across four intensive longitudinal datasets, five candidate predictors were evaluated: three network-structure variables (density, global strength, and maximum modularity coefficient) and two network-informed variables based on node centrality, alongside the average sum score of all nodes. Network-structure variables consistently failed to reach significance as predictors of psychological outcomes, whereas an individual's average value of the most frequently central node in the network emerged as the most consistent predictor across datasets.
Sources: Johal & Rhemtulla (2024)
Related Terms
- psychometric networks (1 shared article)
- intensive longitudinal data (1 shared article)
Applications
Prediction and Centrality
Centrality coefficients identify which nodes are most influential within a network based on their connections to other nodes, and this positioning has been used to motivate their selection as predictors of external psychological variables. In clinical contexts, the value of the most central node in eating disorder symptom networks has been associated with treatment response for anorexia nervosa, level of clinical impairment, and comorbid depression, while the most central symptoms in obsessive-compulsive disorder networks predicted participants' anxiety and depression scores.
Sources: Johal & Rhemtulla (2024)
Prediction and Network Density
Network density, the proportion of non-zero edges present among all possible edges in a symptom network, has been examined as a predictor of clinical outcomes across patient groups. Studies comparing patients who persisted in a major depressive disorder diagnosis after treatment with those who did not found higher density in the former group, yet density failed to function as a significant predictor of psychological outcomes when evaluated as an individual-level summary variable in intensive longitudinal designs.
Sources: Johal & Rhemtulla (2024)
