Augmenting visual analysis in single-case research with hierarchical linear modeling

Area of Science:

• Educational Psychology
• Applied Behavior Analysis
• Quantitative Research Methods

Background:

• Single-case research designs (SCR) are crucial for evaluating interventions.
• Visual analysis is the traditional method for interpreting SCR data.
• Limitations exist in visual analysis for complex designs and detecting subtle individual differences.

Purpose of the Study:

• To demonstrate how hierarchical linear modeling (HLM) can enhance visual analysis in SCR.
• To quantify group-level effects and individual variability in SCR data.
• To highlight the combined utility of HLM and visual analysis for identifying evidence-based interventions.

Main Methods:

• Growth modeling using HLM was applied to a delayed multiple baseline design with an embedded changing criterion design.
• Repeated-measures HLM and visual analysis were used with simulated data from an ABAB design.
• Data involved students with moderate intellectual disabilities (MoID) in a literacy project.

Main Results:

• HLM quantified group-level functional relations and revealed significant variability in baseline probes and growth trajectories.
• Receptive vocabulary and print knowledge were significant predictors of sight-word acquisition and growth rates.
• HLM confirmed functional relations in simulated data and identified significant participant-level variance undetectable by visual analysis alone.

Conclusions:

• Combining HLM with visual analysis offers a more robust approach to SCR data interpretation.
• HLM provides quantitative insights into individual differences and intervention effects.
• This integrated methodology supports the identification and validation of evidence-based interventions.