Each plot, (A)-(C), below shows the relationship between a b…
Each plot, (A)-(C), below shows the relationship between a baseline covariate (e.g., pretest) on the horizontal axis and an outcome variable on the vertical axis, with points labeled by treatment group. Treated cases are represented by blue triangles and the control cases are represented by green circles. Assume treatment effects are constant. For each item below, indicate the plots for which the statement is true. For example, for the first item below, if you decide that A and C are not balanced at baseline you would choose letters ‘A’ and ‘C’ only (i.e., not ‘B’). The groups are NOT balanced at baseline_______
Read DetailsIn a two-group study, student participants were randomly ass…
In a two-group study, student participants were randomly assigned to either a new math curriculum or business-as-usual control. The researcher measured math achievement at baseline and then again after two months (the primary outcome). The mean score for all participants at baseline was 20. The mean score for all participants at the two month endpoint was 26. The researcher fit an ANCOVA model controlling for baseline score. The prediction equation shown below, where Gi represents group (0 = control, 1 = treated), and Xi represents baseline score. According to the fitted model, what is the adjusted mean of the treated group? \(\hat{Y}_i = 3.5 + 4.6G_i + 0.7X_i\)
Read DetailsSuppose interest centers on a test of the mean of the first…
Suppose interest centers on a test of the mean of the first group against the average of the means of the other two groups. If the researcher defines a contrast ψ = c1µ1 +c2µ2 + c3µ3 and tests H0 : ψ = 0, the values c1 = 1, c2 = −1, c3 = −1 correspond with a test of the desired hypothesis.
Read Details