View in EDS

Synthesis of Single-Case Design Mediation Effects Using Two-Stage Multilevel Modeling

Saved in:
Bibliographic Details
Title: Synthesis of Single-Case Design Mediation Effects Using Two-Stage Multilevel Modeling
Language: English
Authors: Mariola Moeyaert, Milica Miocevic (ORCID 0000-0001-8487-3666), Yaosheng Lou, Matthew J. Valente (ORCID 0000-0001-9130-2255)
Source: Grantee Submission. 2026.
Peer Reviewed: Y
Page Count: 53
Publication Date: 2026
Sponsoring Agency: Institute of Education Sciences (ED)
Contract Number: R305D240044
Document Type: Reports - Research
Descriptors: Research Design, Intervention, Hierarchical Linear Modeling, Monte Carlo Methods, Statistical Inference, Statistical Analysis, Statistical Bias, Error Patterns
Abstract: Mediation analysis in Single Case Experimental Designs (SCEDs) allows for evaluating mechanisms through which interventions achieve effects for a single individual. Modeling approaches have been described and empirically validated for mediation analysis in the AB phase design (i.e., SCED involving only one participant with one baseline and one intervention condition). However, no study to date has focused on synthesizing indirect effects across participants from a multiple baseline design (MBD). The current study fills this gap by investigating the performance of the two-stage multilevel modeling approach to synthesize indirect effects using a large-scale Monte Carlo simulation study. An empirical demonstration with interpretation of results is provided. The results are promising for estimation of the indirect effect as unbiased effects are obtained under all conditions that have more than three participants. If statistical inference is of interest, then the approach can be recommended for at least 20 study participants and a non-zero value for the mediator-outcome relation. Under these conditions, the coverage proportion is close to the nominal level of 0.95 and the Type I error rate is controlled. To obtain sufficient power to identify a true indirect effect, at least eight participants, a low between-case variance in mediator-outcome relation and a mediator-outcome relation of 0.39 or higher is needed. [This paper will be published in the "Behavior Research Methods."]
Abstractor: As Provided
IES Funded: Yes
Entry Date: 2026
Access URL: https://link.springer.com/journal/13428
Accession Number: ED680531
Database: ERIC
Description
Abstract:Mediation analysis in Single Case Experimental Designs (SCEDs) allows for evaluating mechanisms through which interventions achieve effects for a single individual. Modeling approaches have been described and empirically validated for mediation analysis in the AB phase design (i.e., SCED involving only one participant with one baseline and one intervention condition). However, no study to date has focused on synthesizing indirect effects across participants from a multiple baseline design (MBD). The current study fills this gap by investigating the performance of the two-stage multilevel modeling approach to synthesize indirect effects using a large-scale Monte Carlo simulation study. An empirical demonstration with interpretation of results is provided. The results are promising for estimation of the indirect effect as unbiased effects are obtained under all conditions that have more than three participants. If statistical inference is of interest, then the approach can be recommended for at least 20 study participants and a non-zero value for the mediator-outcome relation. Under these conditions, the coverage proportion is close to the nominal level of 0.95 and the Type I error rate is controlled. To obtain sufficient power to identify a true indirect effect, at least eight participants, a low between-case variance in mediator-outcome relation and a mediator-outcome relation of 0.39 or higher is needed. [This paper will be published in the "Behavior Research Methods."]