ABSTRACT

One of the most important concepts in measurement is reliability. Reliability indicates whether a measurement device can measure the same characteristic over and over and get the same results. This chapter looks at a theoretical model of reliability, at different methods that is used to estimate it, and what to do in classrooms to improve the reliability of tests. Classical measurement theory uses the concept of variance to explain reliability. True scores are the scores that students should get if everything worked perfectly. Subject effects are all the sources of error related to the individual—any personal issues that cause the person to score other than what he/she should score. These can include, among others, illness, medication, excessive sleepiness, anxiety, personal preoccupations, and luck. Test effects are the various issues related to the test itself that cause the student to score differently than he/she should.