Suppose the weight recorded using the scale is 987g. When you get home, you weigh the sugar using the weighing scale in your kitchen. You buy a packet of 1 kg sugar from the grocery store. Similarly, a high degree of error results in a measurement of low reliability. A low degree of error results in a measurement of high reliability. The degree of error and reliability share an inverse relationship.

It is the degree of error that decides the reliability of a measurement process. Error is the difference between the measured value of a variable and the actual value of the variable. But what does a constant measurement procedure mean?Īll measurement procedures are associated with an error. For this to be realized, the measurement procedures that measure a constant variable must themselves be constant to produce the same outcome. The concept of reliability is centered on the assumption that the feature or variable being measured remains constant and stable. However, if there is a significant difference between the scores, the IQ test cannot be considered representative of X’s intelligence and is therefore unreliable. If this is the case, then the IQ test can be considered to be reliable. Since it is unlikely that the individual’s intelligence would change over a week, the scores obtained by X in the IQ test at the beginning and end of the week should be similar. The questions of the IQ test should be designed in such a way that they solely measure the intelligence of the individual X and not other features such as reading comprehension or confidence level. For example,Ī researcher wants to measure the intelligence of a person X. Such instruments are considered reliable when they can repeatedly obtain a similar measurement when used again and again. And, these measurement tools are not limited to the scientific instruments used in laboratories the reliability of questionnaires, surveys, and interview questions which are also a part of the measurement process, is equally important. This “measurement process” may be a specific measure, tool, test, or experiment. In the overall field of science, reliability is a way of evaluating the quality of the measurement process used to collect data from the world around. In science, reliability is also viewed as a measure of consistency, but its definition is slightly restricted. Therefore, the evidence that forms the basis of any particular field of science needs to be reliable. Science is based on understanding the natural world around us, based available on evidence.

Test of reliability in research