Measurement and timing
Review whether waves share a vertically linked IRT theta metric, and assess implications of using raw or standardized scores. Compare nominal wave spacing with actual assessment dates.
CATCH IT BEFORE REVIEWER 2 DOES
Latent Growth & Growth Mixture Modeling for Longitudinal Cohort and Intervention Studies
Submission-ready trajectory modeling, scale-comparability diagnostics, and multi-start mixture sensitivity for research teams with 3 to 8 measurement waves.
For PIs, Co-PIs, research directors, and analysis teams working with ECLS-K, ECHO, HRS, MCS, or longitudinal trials funded by IES or NIH.
Who this is for
- Education cohort researchers
- Child-health developmental teams
- Aging and panel investigators
- Longitudinal intervention evaluators
- Data centers seeking an independent methods audit before release
What we audit and deliver
Missingness and attrition
Describe wave gaps and baseline imbalances. Fit FIML growth models and examine a plausible departure from MAR.
Growth and survey design
Compare unconditional and conditional growth. The kit demonstrates linear and quadratic forms and school-clustered, weighted inference. Applied scopes can include piecewise or free-loading forms.
Mixture sensitivity
Compare one through three classes across starts and residual structures. Use posterior shares, entropy, BIC, and replication as contextual evidence when deciding whether to retain continuous growth.
The kit includes the synthetic audit report in HTML and Word, a 12-point audit checklist, the Project Scoping and Fee Guide in HTML, and four R scripts with their Quarto template.
- eclsk_lgm_gmm_demo_report.html
- eclsk_lgm_gmm_demo_report.docx
- longitudinal_audit_checklist.md
- project_scoping_and_fee_guide.html
- 01_simulate_eclsk_reading.R
- 02_fit_lgm_lavaan.R
- 03_fit_gmm_sensitivity.R
- 04_export_diagnostics.R
- report_template.qmd
REPRODUCIBLE DEMONSTRATION
Six waves of synthetic reading trajectories
Explore achieved results from simulated data patterned on ECLS-K:2011. Fit measures and mixture comparisons are contextual screening guides.
Loading achieved diagnostics…
Quadratic latent growth model
| Metric | Achieved |
|---|
Exploratory mixture sensitivity
| Classes | BIC | Entropy |
|---|
Principal methodologist
Nedim Yel, Ph.D. is the founder and principal statistician at Data Analysis & Statistical Solutions and the creator of Reviewer Ready and StatPower. He holds a Ph.D. in Measurement, Statistics, and Methodological Studies and has 15+ years of experience in quantitative methodology, psychometrics, longitudinal multilevel and structural equation modeling, and IES- and NSF-funded research.
Request a scoped longitudinal pilot
Tell us about your study and decision timeline. Please share study metadata and approximate counts only. See our privacy policy.