Reliability

Reliability is the consistency of a measurement or procedure. A reliable instrument produces similar results when repeated under similar conditions. Common forms include test-retest reliability (stability over time), internal consistency (whether items on a scale measure the same construct, often assessed with Cronbach's alpha), and inter-rater reliability (agreement between different coders or observers).

Reliability is necessary but not sufficient for validity. A scale that is consistently two kilograms off is reliable but not valid; an unreliable measure, however, cannot be valid, because its scores are dominated by noise.

For a thesis, reliability must be evidenced, not asserted. Report the appropriate coefficient for each scale, describe how coders were trained and agreement was checked in qualitative work, and standardize procedures so results do not depend on who collected the data. These details are exactly what examiners look for in the methodology chapter.