The percentage of non-overlapping data (PND) is a statistical metric used in single-subject research to summarize the effectiveness of an intervention. It is calculated by identifying the percentage of responses in a treatment condition that are more extreme than the single most extreme response recorded in the corresponding control or baseline condition. For example, in a study by Hall and his colleagues, all measurements of Robbie's study time during the first treatment condition were greater than the single highest measurement recorded during his initial baseline phase. This resulted in a PND of $100\%$, indicating a very strong treatment effect. Generally, a higher percentage of non-overlapping data signifies a more robust treatment effect.