Skip to content

Latest commit

 

History

History
105 lines (76 loc) · 9.99 KB

README-old.md

File metadata and controls

105 lines (76 loc) · 9.99 KB

logl Quick links: news - projects - tools - log - participants - dashboard - GSA poster

#Concurrent decline of body and mind

IALSA-2015-Portland workshop performs a coordinated analysis with replication (CAR) of decline in physical and cognitive functioning during later adulthood.

The following fou4 graphs cross-tab the counts of submitted models with Cognitive Measures:

model space 5D
The numbers in the cells count models of that specification. Each color adds up to 100%. "univar" indicates univariate growth models fitting either cognitive or physical process. Submitted models vary in respect to their:

  • study_name - (e.g. eas, satsa) in which it was estimated
  • model_number: (u0, u1, u2, b0, b1, b2) alphanumeric of model form (u,b) and shape (int0rcept, s1ope, qua2radic )
  • subgroup: (female, male) subsets on which models are estimated
  • model_type: (0, a, ae, aeh, aehplus, full), a.k.a. CovSet, Predictor Set
  • physical_measure - (e.g. grip, gait, pek)
  • cognitive_measure - (e.g. block, digits, animals)

Any model counted in the above graph can be specified with a form nested within the general specification:
general_model_specification

The covariance structure will be given as general_model_specification
For details see model specification.

Tools

Analytic tools available to the workshop participants include dynamic reports and interactive apps. Both reports and apps are regenerated with each new influx of model submissions.

Reports

  • Rename and collapse gives account of how model filename elements (subgroup, model_type, cognitive_measure, etc.) are corrected, renamed, and reclassified. Incorporates the results of Teleconference 1 on Jun 12.
  • Model Counts and Model Essentials give basic tabled view of data in model space.
  • Current status update for each study: EAS, ELSA, HABC, ILSE, NAS , NuAge, OCTO-Twin, RADC, SATSA
  • model_space graphing how the measures are mapped into domains. We need to understand how best to organize our models for the analysis, in which model (not individual) is the primary unit .
  • reproduce script that re-generates all these reports.
  • scripts README overview scripts that replicates data preparation

Apps

Participants

Study Contact
EAS measures
ELSA Annie Robitaille
HABC Chenkai Wu
ILSE Philipp Handschuh
NAS Lewina Lee
NuAge Valerie Jarry
OCTO-Twin Marcus Praetorius
RADC Cassandra Brown
SATSA Deborah Finkel

Contact IALSA team

Background

The current repository analyzes the results of the IALSA Analysis Workshop (pdf overview) that was held in Portland, Oregon from Monday February 23rd (8:30am) to Wednesday February 25th (12:30pm), 2015.

The primary aim of the workshop was to examine associations between changes in physical functioning (i.e., grip strength, pulmonary function, chair stands, walking speed) and cognitive functioning (i.e., measures of speed, memory, reasoning, executive functioning) in multiple-study comparative framework. Bivariate growth processes were estimated to evaluate dynamic associations in change and variation within and across these domains and in relation to sample and individual differences in age, lifestyle, and health outcomes. The results from these analyses will extend recent systematic reviews and meta-analyses (e.g., Clouston et al., 2012, Epidemiological Reviews and will be submitted for publication as a set of independent brief reports, capped by a literature review/overview of analysis and completed with a research synthesis summary.

Projects

See projects page for a detailed overview of the projects planned before the Portland conference.

At the time a project shapes up it gets its own repository, which will eventually host the publication manuscript and all contingencies for reproducibility. Currently, the following repos are gearing toward producing manuscript for submission:

  • Portland-physical-cognitive repo follow the physical-cognitive track, evaluating the longitudinal models in which physical and cognitive outcomes form a bivariate linear structure.

News

Log

The agenda and outcomes of each teleconference following the Portland workshop


Quick links: news - projects - tools - log - participants - dashboard

edit me