Fig. 2From: Does group-based trajectory modeling estimate spurious trajectories?Individual trajectories of difficulty initiating and maintaining sleep for identified trajectories (left panel) and a sample of participants (right panel), NDIT Survey cycles 1–20Back to article page