Measuring procedural fidelity in behavioural research.

Procedural fidelity (PF) refers to the implementation of a research plan, as intended. Measuring PF should be done not to satisfy minimum requirements for publication but to provide useful information to implementers, researchers, and consumers of research. Measurement of PF requires careful consideration and planning, and should include: naming important steps to be completed in each condition, defining each step, choosing a measurement system, determining which analysis is beneficial, training observers, conducting observations, analysing fidelity and identifying discrepancies, and reporting adequate data for consumers. Careful measurement and reporting of PF data will lead to advances in determining which interventions are effective, for whom, and under what conditions.