2016-08-16-Tuesday-TC-Lookup-Function

[andkov]

@ampiccinin , @GracielaMuniz , @wibeasley

Agenda

Updates

• 1. New function is available to look-up results of individual models.
• 2. Data flow for the secondary analysis is conceptualized with regards to tables.

     Decisions

• 1. The prototype includes covariances between : 1(Intercept Physical - Intercept Cognitive), 2(Slope Physical - Slope Cognitive), 3(Intercept Physical - Slope Cognitive), and 4(Intercept Cognitive - Slope Physical). However, it seems to me that the latter two should be replaced by 3(Intercept Physical - Slope Physical) and 4(Intercept Cognitive - Slope Cognitive) as more descriptive of the model. What do you think?
• 2. Add parameter count?
• 3. Any new ideas what to do with estimated vs computed correlations? They mostly align, but they are not identical. Using computed will require some non-trivial adjustment to the script because they do not have the standard est-se-wald-pval form provided by MPlus on which much manipulation relies.

[andkov]

Decision

• 1. Add both covariances. The table should include 6 covariance components. Consider adding ALL covariance components, so that people can decide for themselves what to interpret (laden with danger of inattantion)
• 3. Prioritize requesting additional model running by original drivers. Prefer calculated, because it offers greater control.

[andkov]

• GSA on Nov 16-20, 2016. Many original drivers will be attending. Think about preparing a tutorial to show how the available results can be used to foster publication production.
  • Using GSA to experiment
  • Offer the specific rules for interpretation?
  • Should workshops be the first step in a 1 or 2-year long collaboration?

GSA interpretation workshop

• the deadline to announce the workshop: Sep 16
• We scheduled a brainstorming session for 2016-08-30-08:00 PDT