Statistics Research » Statistical Design & Analysis Page Tools AddThis

Longitudinal Analysis

Longitudinal analysis involves studying series of observations obtained from respondents over time and is often used in clinical trials, panel surveys, laboratory experiments, and genetic studies among others. Multilevel modeling takes into account correlated observations due to sampling design, stratification, clustering, and repeated measures. Various data analysis methods can handle continuous, dichotomous, ordinal, and nominal outcomes.

Focus Areas

  • Survey: account for survey design, sampling units, stratification
  • Clinical trials: adjust for clustering or stratification of participants
  • Longitudinal cohort or panel studies: account for repeated measures on the same subject
  • Cluster randomized trials: control for randomization of groups instead of individuals
  • Genetic analysis: adjust for repeated tests or observations

Methodologies/Techniques

  • Generalized estimating equations (GEE)
  • Generalized linear mixed models
  • Non-linear mixed models
  • Structural equation modeling (SEM)
  • Growth mixture modeling (GMM)
  • Survival analysis
  • Weighted least square (WLS)
  • Mantel-Haenszel strategy for repeated measures

Applications

  • Analysis of development over time
  • Multisite longitudinal clinical trials to evaluate the effect of a therapy
  • Multisite cluster randomized trials to evaluate an intervention effect
  • Genetic analysis (for example, localization of genes influencing common complex diseases and their risk factors)

Projects

  • Study of Early Child Care and Youth Development (SECCYD). Naturalistic multisite longitudinal examination of normative child development from birth through age 15.
  • Data Coordinating Center (DCC) for the NICHD Neonatal Research Network (NRN). Multicenter observational and clinical trials of very low birth weight infants, which sometimes involve stratification by center or gestational age, or cluster randomization by family or hospital.
  • Data Coordinating Center for the NIH-DC Initiative (DCI). Development and evaluation of interventions that may reduce infant mortality in Washington, DC. Studies often involve longitudinal follow-up of participants.
  • The Data and Coordinating Center for Global Network for Women's and Children's Health (GN). Innovative and flexible research network that is responsive to the most critical existing and emerging health needs and public health problems of women and children globally. Studies often involve cluster randomized trials and longitudinal follow-up of participants.
  • Maternal Lifestyles Study (MLS). Longitudinal cohort study comparing infants exposed to cocaine or opiates to nonexposed infants.

Contact us for more information

Related Content