Reliability

Reliability is the consistency of a measure — whether it produces stable results across items, occasions and raters. A reliable instrument is not necessarily valid: a scale that consistently measures the wrong thing is reliable but useless. Reliability is necessary for validity but does not guarantee it.

It comes in several forms. Internal consistency asks whether items of a scale hang together (Cronbach's alpha, composite reliability). Test-retest reliability asks whether the measure gives similar results on repeated administration. Inter-rater reliability asks whether independent coders agree, quantified with Cohen's kappa or intraclass correlation.

The working threshold for internal consistency is 0.70, with 0.80 and above considered good and values above 0.95 suggesting redundant items. Report reliability separately for each construct rather than for the whole questionnaire, and calculate it on your own sample rather than citing the original authors' figure.