Reliability of Psychology Test (Coursework Sample)

Reliability of Psychology Test Name Institution Reliability Test quality is a critical part of psychological testing, one of the most important elements used when gauging quality is reliability. It is concerned with consistency, dependability and reproducibility Thambirajah (2005). Reliability in psychology testing is the degree to which the outcome of the test remains unaffected by irrelevant variations and procedures of testing. Reliability shows the extent to which the test obtained by exterminating a person would remain the same if the person is subjected to the same test again on a different occasion. In a narrow perspective, it is the accuracy in which any psychological test measures a given characteristics of interest. It is the extent that the test score is error free. Charles Spearman was the first to explain that measurements are always open to error, and scores on a given test varies from time to time and from one circumstance to another, he called it true of score theory, commonly known in psychology as the classical theory. It states that the observed score X is the sum of true score T and the error component E (Thambirajah 2005). X= T+E True score T is the ideal measurement; it is what the examiner strives to achieve; thus it is constant, the error component E is based on the assumption of random distribution and cannot be eliminated thus a variable. The sources of the error are varied for example, the degree of stress and variation in condition of the test. Because of the assumption that E is randomly distributed, there is a need to calculate the reliability coefficient for better interpretation. For example, a reliability coefficient of 0.8 can be interpreted as, 80% of the variability test score can be accounted for by true score variance and 20% accounted for by error score variance. The reliability of the test is achieved by obtaining two measurements and conducting a comparison between the two. Reliability minimum permissible figure is 0.8, below that score the test is rendered meaningless. Most psychometric test reliability coefficient is at 0.9 0r above. There are several methods used for in computing test reliability. This paper will discuss Test Retest reliability (Thambirajah, 2005). Test Retest reliability Estimation of Test Retest reliability involves administration of a test to a single group of examinees on two separate occasions. The time span between the first administration and the second one is approximately three months Gabrienya (2003). The relationship between the examinees, different scores, are then estimated using statistical correlation to determine the similarity of the test; the closer the two results the greater the reliability of the test. Good test should be stable over separate administrations. The test is affected by factors like fatigue, anxiety, boredom, guessing, poor test instruction and practice effect (Gabrienya, 2003). Impact of Test Retest reliability The Test Retest reliability if used on variables that are stable and are not prone to change such as IQ test, yield very reliable result, but when the test is applied on psychological characteristics that vary the outcome is likely not to be reliable because of recall bias, learning and true changes. The test would give a false impression of reliability upon evaluation but in reality, it is not reliable because the examinees may have been influenced by factors such as recall bias and guesswork (Qu, 2010). Reliability strategy The main aim of carrying out a test is to get a reliable outcome, which will aid in making vital decisions. Most reliability techniques are affected by other factors that lower the level of reliability. In order to increase the reliability level, the examiner should use greater standardization met...