Reliability and Validity

Measurement is a method through which concepts which cannot be observed are converted or linked to events which can be observed. Like emotion which is abstract is linked to questionnaire which help quantify the subject. Conceptual definition is expression of the main idea behind concept. This definition helps differentiate the study concept from other similar and confusing concepts. An example of concept definition is clear mentioning of the happiness the viewer feels whenever he watches certain advertisement. Operational definition is translation of the main concept of interest which is mostly unobservable into observable events by clearly explaining the important observable characteristic of the concept. Operational definition is important because it behaves as a link between the theory and data collection in reality. Even when conceptual definition is same for different researchers or people still it has been seen that observation event is separate for most. The operational definition brings clarity in the translation part of converting from unobservable concept to observable event. It is a necessary condition that measurement should be reliable. Reliability here means that when a test is conducted several times then it should be free from random error and should give consistent and stable result. Higher is the similarity between the answers, higher will be the reliability. Two main methods of measuring the reliability of the measurement are as follows: Test-retest: In this method the test is repeated again and again by keeping the conditions of the test same. For high reliability the results obtained from each test should be same or with negligible difference. Alternative form: In this method, different test which are made for the same purpose are administered and then reliability is accessed by the determining the difference between the results. It is a necessary condition that measurement should be valid. Validity here means that when a test is conducted then it is seen that results of the test are actually measuring the concept which is tried to be measured. Sometimes these tests are such that the result given are different than the operational definition of the concept. Then it is said to have low validity.