Cross-lagged panel analysis involves the examination of causal influences between variables over time. "Crossed" refers to the fact that reciprocal relationships between variables are estimated. "Lagged" refers to the fact that these relationships are estimated at multiple points in time. "Panel" refers to the use of longitudinal data in which observations are recorded at multiple time points.