This m-file deals with repeated measurements at t time points obtained from s groups of subjects or multiple samples. With or without same size. The simplest model is:
x_hij = m + g_h +t_j + (gt)_hj + p_i(h) + e_hij
where: x_hij = response at time j from the ith subject in group h; m = overall mean; g_h = fixed effect of group h; t_j = fixed effect of time j; (gt)_hj = fixed effect for the interaction of the hth group with the jth time; p_i(h) = random effects for the ith subject in the hth group, and e_hij = independent random error terms.
For we are interested test for differences among groups, test differences among time points, and test the significance of the group x time interaction. The first test requires the assumption that the within-group covariance matrices are equal. The two last-ones tests require this same assumption and that the sphericity condition is satisfied. In general, these assumptions are required for all tests of within-subjects effects (repeated measures).
Inputs:
X - data matrix (Size of matrix must be n-by-4;dependent variable=column 1;group=column 2;independent variable=column 3;subject=column 4).
alpha - significance level (default = 0.05).
Output:
Complete Analysis of Variance Table. |
