Do descriptors influence eyewitness performance and lineup fairness?

Eyewitness descriptions play a crucial role in investigations, as they have a significant influence on investigation leads and lineup construction. However, little attention has been given to how descriptions are compiled into a single descriptive profile (i.e. descriptor techniques) in research and practice when there are multiple eyewitnesses. We examined how different descriptor techniques – comprehensive (all non-contradictory and contradiction-resolved descriptions), random (a random description subset), and consensus (commonly-provided descriptions) – influence eyewitness performance and lineup fairness. We expected lineups constructed with comprehensive descriptors to yield the highest fairness and lowest discriminability, as they would have the highest suspect-filler similarity due to inclusion of more details. Participants (N = 363) watched two mock-crime videos, completed a distractor task, made a lineup decision, and rated their confidence. Each participant completed one target-present and one target-absent lineup, followed by attention-check and demographic questions. Lineups constructed with comprehensive descriptors had the highest discriminability but the lowest fairness, while lineups constructed with random and consensus descriptors were fairer but showed reduced discriminability. Discriminability only differed significantly between the lineups constructed with comprehensive and random descriptors. Therefore, the compilation of descriptions into descriptors may have an impact on lineup performance and fairness – with both methodological and practical implications.