[Determining the effectiveness of therapeutic interventions in a time series of examination plans]

Some typical effects of therapeutic intervention in time series designs are considered and suggested for heuristic evaluation by Kendall-tests. The pre-intervention phase is assumed to be stationary as to an indicator variable and the post intervention phase is assumed to change in level or direction. According to the effects assumed, anchor series of ranks are constructed and compared to rank-transformed time series observations. Large Kendall-scores indicate agreement between assumed and observed series of ranks, i. e. either change in level and/or in direction.