# convert condition labels also to factors, relabel factor levels
d <- cbind(d, colsplit(d$EXP, names=c("Cue", "Target")))
d$Cue <- relevel(d$Cue, 2)
d$Target <- relevel(d$Target, 2)

# add variables with usefule contrasts 
# ... for C-T combinations
# ... ... treatment (EXP=EXP0; default)
d$EXP0 <- C(d$EXP, matrix(c(  0,  1, 0, 0,   # (1)  VA vs VV
                              0,  0, 1, 0,   # (2)  AV vs VV
                              0,  0, 0, 1    # (3)  AA vs VV
                                          ), 4, 3), 3)  
contrasts(d$EXP0)
# ... ... nested
d$EXP1 <- C(d$EXP, matrix(c(-.5, +.5,     0,    0,   # (1)  type of target |visual cue
                              0,   0,    -.5, +.5,   # (2)  type of target |auditory cue
                            -.5, -.5,    +.5, +.5    # (3)  type of cue  
                                                 ), 4, 3), 3)  
contrasts(d$EXP1)

# ... ... crossed
d$EXP2 <- C(d$EXP, matrix(c(-.5, -.5,   +.5,   +.5,   # (1)  type of cue
                            -.5, +.5,   -.5,   +.5,   # (2)  type of target
                            +.5, -.5,   -.5,   +.5    # (3)  cue x target 
                                                   ), 4, 3), 3)  
contrasts(d$EXP2)

# ... for number of microsaccades--see contr.sdif(3) of MASS package
# ... ... repeated
#d$MN0 <- C(d$MN, matrix(c(  -2/3, +1/3, +1/3,   # (1)  0 vs. 1
#                            -1/3, -1/3, +2/3    # (2)  1 vs. 2
#                                     ), 3, 2), 2)  
#contrasts(d$MN0)

# ... ... Helmert
#d$MN1 <- C(d$MN, matrix(c(-1, +.5, +.5,  # (1)  0 vs. 1, 2+
#                           0, -.5, +.5   # (2)  1 vs. 2+
#                                      ), 3, 2), 2)  
#contrasts(d$MN1)