Working Paper No. 2005 | 02
Temporality and Intervention Effects: Latent Trajectory Analysis of a Homeless Mental Health Program
by William McAllister (ISERP), Li Kuang (Public Health), and Daniel Herman (Public Health)
Intervention analyses which incorporate temporality over a followup period typically note differences in the patterns of "single-curves" for each the experimental and control groups or differences in temporally-based taxonomies between experimentals and controls. But the former fails to allow for the possibility of subgroups of multiple trajectories and the latter collapses time (e.g., average spell durations) and arbitrarily creates cut-points to form its taxonomies. This paper investigates the utility for intervention research of using latent class growth analysis (LCGA). This method incorporates the more complete temporal information used by single-curve approaches to statistically identify the multiple subgroups at the heart of the taxonomic approach. The authors do this by reanalyzing a critical time intervention study (CTI) of homeless mentally ill men that used both single-curve and taxonomic approaches. By finding, among other things, differences between experimentals and controls in the number, sizes and patterns of latent subgroups than were found in the prior analysis, this paper suggests the utility of LCGA for assessing service interventions.
Download 
See Also
- Research grant: Temporal Dynamics, Intervention Analysis and Policymaking: Latent Trajectory Group Analyses of Service Interventions for Homeless Mentally Ill
- Working paper: Policymaking and Caseload Dynamics: Homeless Shelters
