A multiple-sequence variant of the multiple-baseline design: A strategy for analysis of sequence effects and treatment comparison

The design logic and potential uses of a variant of the multiple-baseline d esign that implements multiple treatment sequences is described. The multip le-baseline multiple-sequence (MBL-MS) consists of two or more multiple-bas eline designs that are interlaced with one another and include all possible sequences of the treatments. The counterbalanced sequencing of treatments permits observation of any sequence effects that are present. This sequenci ng can permit identification of an optimal treatment sequence and compariso n of relative efficacy of treatments. This variation of the multiple baseli ne also extends the available treatment comparison designs by providing an additional design that does not require reversal of treatment or sequence e ffects. Because of vulnerability to ceiling and floor effects, the MEL-MS d esign appears to be primarily useful for the comparison of treatments that take several sessions to produce an optimal Level of responding or when the behavior is not subject to floor or ceiling effects.