From Data to Decisions: Applying Inferential Statistics to Longitudinal N-of-1 Data to Evaluate Intervention Effects – A Narrative Review

Introduction: Psychology is shifting from group-based designs to idiographic approaches, raising questions about suitable analytical methods. While a full review across subfields is beyond this paper’s scope, psychotherapy research serves as a representative example. Single Case Experimental Designs (SCEDs) can be used to estimate meaningful intervention effects for individuals, but selecting among many evaluation methods is challenging. This study provides a comprehensive overview of SCED analysis methods, assessing their strengths and limitations to guide scientific and clinical decision-making. Method: A literature search was conducted to identify methods suited to compare a baseline to an intervention phase, to be used with idiographic or non-standardized items and with repeated evaluations during ongoing therapy. To illustrate these methods, data from a clinical single case is analyzed and interpreted as an example. Results: Seventeen relevant peer-reviewed articles on SCEDs and clinical decision-making were identified. NAP, Tau-BC, and the Bayes Factor emerged as particularly suitable methods, though further validation is needed. Method selection highly depends on data characteristics, with nonoverlap approaches preferred for non-parametric data and SMA or DARCI suitable for small datasets. The case study revealed challenges in interpreting and combining different methods, emphasizing the need for clearer guidelines. Discussion: While no single method is universally optimal, future research should refine existing approaches, compare them systematically, and integrate clinical perspectives to develop clearer guidelines. Beyond identifying a gold standard, ensuring practical implementation remains a key challenge.