Beyond classical content validity indices: methodological aspects and reporting recommendations for nursing research

Expert panels are widely used in nursing research to judge whether items, definitions or diagnostic components adequately represent a content domain. However, numerical indices for these tasks are often treated as if they all measured the same entity.

To examine classical and less commonly discussed procedures used to summarise expert judgement in content-validation studies.

A narrative methodological review and analytical demonstration used seminal and later methodological sources to address five predefined questions; hypothetical datasets were used only to reproduce and compare the procedures.

The procedures are not interchangeable; while some summarise the proportion of experts endorsing relevance or essentialness, others preserve graded ratings, some adjust a proportion for a chance model and others focus on disagreement, coverage or item-to-domain mapping. Because these procedures transform expert judgements differently, the same ratings can produce different numerical values. Small panels can also produce imprecise estimates even when a point estimate appears favourable.

Researchers should choose a procedure because it answers their methodological question. Clear reporting of the target domain, expert task, original ratings, formula, panel size, uncertainty and complementary agreement or reliability information makes content-validation evidence easier to interpret and use.