Reliability and Validity Paper Essay Example | Topics and Well Written Essays - 750 words

Reliability and Validity Paper - Essay Example † and then proceeds with â€Å"How it will be measured?† (Desarbo, William, & Babin, 2009) A measurement is defined as ‘good’ if it fulfills the criteria of validity and reliability. Reliability: Reliability measures the extent to which the test yields consistent results each time it is conducted and across various situations and includes the following (Neukrug, 2008). Test-retest: This involves conducting the same scale or test to similar respondents at two different points in time (Neukrug, 2008). For example, a researcher interested in measuring the number of times an average adult worker falls sick within a year discovers that 68% of the population falls sick at least thrice. If this test is conducted again, under similar conditions, and it is discovered again that 68% of the population falls sick at least thrice a year, it indicates that the test is reliable. Split-half: Under this the researcher checks one half of the results of a test against the othe r half to check for internal consistency (Neukrug, 2008). For example, a researcher who has measured good quality of human services at a hospital in terms of four items ambience, proper equipment, quick procedure and trained staff can split the results by grouping the results of ambience and proper equipment together and quick procedure and trained staff together respectively and then check the Chronbach’s alpha for these two groups. ... If the results obtained through the two tests show high correlation, then the tests are considered to be reliable. Validity: Validity refers to the extent to which the test measures what it is supposed to measure; that is, the extent to which the test meets its objectives (Neukrug, 2008). The following are the types of validity: Face or content validity: This results from the consensus amongst experts that a scale logically appears to be reliable (Neukrug, 2008). For instance, questions such as â€Å"how many children do you have?† asked by a researcher regarding the effectiveness of family planning programs are generally assumed to stand up to face validity. Criterion Validity: If the measure correlates with another measure of the same construct then it is assumed to have criterion validity (Desarbo, William, & Babin, 2009). If the new measure is tested at the same time as the old measure and is shown to be valid it is known as concurrent validity and if the new measure is te sted at a later time and is shown to be valid then it is predictive validity (Desarbo, William, & Babin, 2009). For example, in tests reveal that the effectiveness of the prototype provide correct estimate of the effectiveness of the final medicine 80% of the time, then the test has predictive validity. If, however, the same correlation is established between the final medicine (after it’s manufactured) and its prototype at the same time, then it will have concurrent validity. Construct Validity: This refers to the extent to which the measure confirms a network of hypothesis based on a theory to which the concepts are related (Monette, Sullivan, & DeJong, 2010). Convergent