LI Analysis Training Series

and Feasibility For evaluators or researchers to obtain access to other sites’ GAIN or EBTx data that has been submitted to the GCC or EBT CC, the evaluator must send an abstract proposing the specific analysis to be conducted and the data being requested to the GCC ([email protected]) for feasibility review and recommendations for any changes. This abstract needs to identify the sample to be used (inclusion-exclusion criteria and samples), the questions to be asked or hypotheses (with as explicit link to variables as possible), the time periods (intake only, or which specific follow-ups may also be needed), the type of analyses planned, and the target audience. To assist in abstract creation, evaluators can download the GAIN-I Administration Manual (Dennis, White, Titus, & Unsicker, 2008) and the GAIN-I Scales and Variables file, an electronic encyclopedia of more than 2,000 pages that provides information about the GAIN items and calculated scales, indices, and calculated variables and helps evaluators determine which variables to use. For grantee sites, the Adolescent Program Support Site and the GAIN Program Support Site are specific to individual grant programs and include reports and summaries for each program. To access these websites, you will need to get the logon and password from your grant’s PI. Examples of data request abstracts and a summary of previous data request topics can be found on the FTP Common site. The list of EBTx summary variables is also available upon request. The GAIN Analysis Team and, if applicable, the EBT analysis staff will do a preliminary review of the abstract to make sure that the proposed analyses are feasible in terms of available data, sample size, and variables, and the staff will also comment on any potential methodological problems. Some key issues that can come up in this process include the absence of the requested variable (or any similar item that could be used instead), a variable that is optional and collected only in a subset of sites, the availability of follow-up data, or limited sample size in a subgroup or over time. To start this process or request feedback on the feasibility of an abstract, submit it to [email protected]. Approval Depending on what data will be used and the analysis being proposed, approval of the final abstract is needed from one or more of the following sources:  Principal investigator (PI) approval: Any PI or group of PIs can authorize the use of data they collectively have rights to. With their approval GCC staff will create a file containing their combined data.  Funder/GCC/EBT CC approval: For analyses on general topics (e.g., withdrawal, comorbidity) that do not use data from current programs or that do not identify sites or clients, approval can be requested from the project funder (e.g., the CSAT project officer) or director of the GCC or EBT CC.  Funder/GCC approval/EBT CC approval plus PI approval: For analyses that will identify specific sites or studies, the GCC will also notify the involved sites’ PIs and give them the opportunity to decline use of their data. Active sites are given 2 weeks to review the abstract and decline use of their data if they so desire. Note that some 3 The FTP Common site is located at ftp://data.chestnut.org; username = Common, password = public (both case-sensitive).