OpenMEE brainstorming

Similar interface, but:

• Remove diagnostic
• Change "grp" to "tmt" (or short for treatment).
• Allow users to enter the SE
• Allow import from CSV
• Add log response ratio

Longer-term stuff

• Randomization/bootstrap

6/25/13 Meeting

• Paste with column names (from a gigantic csv with all outcomes i.e. entire dataset) --> user will explain the meaning of each column

• Effect Sizes: Hedges g OR LnR Regression coefficient Own effect

• Bootstrap methods (investigate what R packages exist for this) confidence limits

• Randomization tests significance level

• Graphical output:

• Exploratory Data Analysis Outlier detection (exploration...) Bias detection research bias publication bias

• Data model Study-->(outcome,time point) --> metrics