Mj. Koehler et Jr. Levin, REGULATED RANDOMIZATION - A POTENTIALLY SHARPER ANALYTICAL TOOL FOR THE MULTIPLE-BASE-LINE DESIGN, Psychological methods, 3(2), 1998, pp. 206-217
A dual-randomization procedure, regulated randomization, is proposed f
or behavioral and educational interventions that incorporate the logic
of single-case multiple-baseline designs. The new approach is sharper
conceptually and methodologically than previously developed approache
s in that regulated randomization maintains the basic integrity of the
multiple-baseline design (namely, the systematically staggered introd
uction of the intervention across the experimental units) while being
statistically practicable with fewer units (N < 4). Moreover, previous
ly suggested nonparametric analyses of multiple-baseline data can be s
ubsumed by a general regulated randomization formula, The regulated ra
ndomization approach provides researchers with a flexible analytic too
l that can take into account specific substantive, methodological, and
statistical trade-offs.