What is the difference between multilevel analysis and latent growth modeling?
Please answer in possibly accessible way. Thank you
Witold
They are the same thing because the random intercepts and slopes of the growth model are not directly observable but are unmeasured or latent - they have to be estimated. That is each higher level has its own growth trajectory and you have measured the higher level unit multiple times (growth can include decline when the slope is negative) and you use a modeling technique to extract these underyling latent trajectories
Here is a paper with occasions nested within counties for life expectancy that makes the connection to group based latent trajectory models where the random intercepts and slopes are not continuous (and follow a Normal distribution) but take on amore limited range of values. It may be easier to 'see' the 'latency' here as then you are identifying the latent groups
https://www.researchgate.net/publication/46541418_International_variations_in_life_expectancy_A_spatio-temporal_analysis
This is a really useful paper as it compares the SEM (with its latent factors) and multilevel approach to longitudinal analysis:
Steele, Fiona (2008) Multilevel models for longitudinal data. Journal of the Royal Statistical Society: series A (statistics in society), 171 (1). pp. 5-19. ISSN 0964-1998;DOI: 10.1111/j.1467-985X.2007.00509.x
Download from
http://eprints.lse.ac.uk/52203/1/Steele_multilevel_models_longitudinal_2008.pdf
to my mind the flexibility of the multilevel approach is clear.
Article International variations in life expectancy: A spatio-tempor...
Multilevel Analysis may be understood to refer broadly to the methodology of research questions and data structures that involve more than one type of unit
- Badges
- Science topic
- Similar topics
- Instrumentation Engineering
- Instruments