Randomizing within Blocks

Randomizing Within Blocks

Randomizing within blocks means assigning participants to conditions in such a manner that each condition has a participant assigned to it before the next round of assignments to conditions is performed. The result is that no condition has significantly more participants than the other conditions, which could happen if assignment was done completely randomly. Furthermore, the ordering of conditions makes it less likely that any kind of drift in procedures that might occur during the study would have a differential effect on the groups, because the order of testing would be very similar across groups. Such a drift in procedures might occur, for example, if the people doing the ratings were subtly changing their rating criteria over time.

Within-subjects designs can be thought of as randomizing the assignment of conditions within blocks, except that the random order of assignment represents the order of the conditions in which the single participant will be tested.

This website includes a random number generator program, which is capable of randomizing within blocks automatically for you. You can download this program free for you personal use. Instructions for how to use this program for creating random orders and for randomizing within blocks is included elsewhere on this website.