Fundamental interest of the scientist is in the measurement. Reliability in a measurement implies the ability to obtain replicate values that differ little because the random error that affects them is small.

For example, by conducting three replicate measurements on each blood sample of Ms. X1, X2, X5 and for Mr X3, X4, X6 to obtain the fasting blood cholesterol value for each individual. The closer the replicate values are, the higher the reliability.