Providence

Meetings

2009 Dec 22

• We will calculate time separately for each measure as the time since the first data from that measure was collected.
• We will look at an interaction between foster care and group and time.

2009 Nov 13

• There were five different cohorts, or groups that started the study at the same time. There were some changes to the protocol between these cohorts. We may want to add a term for cohort in the model with a cohort-group interaction.
• Manuel recommends random slopes for clients and therapists.
• The primary analysis will be intent to treat, and the secondary analysis will take into account implementation.
• Implementation
  • There is a variable in the data giving the date when the therapist completed training modules. We can create a binary variable indicating whether they ever completed the module, or make a variable giving the number of modules they completed. We could put this in the model with a group interaction.
  • There are two implementation variables that are constant for each client: one is a percent, and one is a total. Tomasso will let us know the names.
  • Completion of questionnaires as implementation.
• To establish that the quarterly feedback group is a control group, that is, there is no effect of the infrequent feedback, look at whether there is a change in trajectory at 90 days. Ben pointed out that we may not expect this trajectory to be linear.
• Adjust for initial severity and add group-initial severity interaction. For the baseline SFSS score, we will use the SFSS taken at the first therapy session. The reason for this is that the initial assessment SFSS, while an ideal baseline measurement, is not available for all clients.
• Ben and JoAnn will work on 1.1-1.6 and try to develop an analysis plan in about one week.
• JoAnn will compare descriptive statistics with Tomasso to verify that the data import was successful.

• aaMASTER_DATA_FIXES_121409_cb.doc: This file tracks data-related issues that emerge while conducting the evaluation analysis for the Providence Project.

• aaMASTER_DATA_FIXES_121509_cb.doc: This file tracks data issues that emerge during the providence project eval analyses