A reliable test estimates the consistency of the measurement conducted. Reliability means how the results are measured and if they are found the same way each time they are gathered.Validity is the measure of accuracy. It is the determination of whatever method used is consistently giving accurate results. This is more important because if the results are not correct then there is no reason to use them. They will not offer an appropriate conclusion. There are cases where unreliable tests can be valid, for example in the first Gallup Poll conducted, random digit dialing was employed in order to determine the public's opinion of political candidates. However, in the 1930s when this survey was conducted, the majority of people who owned phones were upper class, Caucasian families or just males. Their opinions on the upcoming election and on the candidates were valid but the sampling itself was unreliable because it was not necessarily consistent with the entire country's views and was not a representative sampling of how the entire country felt. It was just a small elite portion who could be reached (Golay 1996).Kirk (1985) defines four methods of establishing validity as case studies, focus groups, surveys, and blind experimental studies. Case studies are non-experimental, descriptive types of studies where an in-depth record is kept by an outside observer. These records can include observations, recording individuals' experiences, biological data, medical records, family history, interviews, and the results of psychological tests. The objective of focus groups is to interview six to ten people at the same time to obtain their opinion, their view, and see their perspective when evaluating a new household product, the media, or even food. The goal is start an open discussion stemming from a few questions in search of a particular answer. Surveys can have various different objectives from determining employee satisfaction or finding out the opinion of individuals about the government. Surveys can be conducted anonymously, without consequence to the person taking the survey or they can be taken to determine information for an individual. Finally there are blind experimental studies, where individuals are not told the objective of the study until after it is completed. There are independent variables, which are used to measure change and a dependent variable, which is the constant results are being compared to. The benefits of experiments are that they help researchers determine cause and effect relationships by manipulating one variable. The similarities between the four methods are that they can be used in any type of situation, they can be used with many different types of people, and each method offers participants complete anonymity. The differences between the four, is the time to takes to conduct each method will vary significantly, the information obtained will be different, and the people who conduct these methods will be different.
3. What are the four methods of establishing reliability How are these methods similar How are they different
There are four methods of testing reliability, as defined by Rudner and Shafer (2001), test-retest reliability, inter-rater reliability, parallel -forms reliability, and internal consistency reliability. In test-retest reliability, a test is administered twice and at two different times, this type of reliability measures whether or not there is a change in the quality of results over time. In inter-rater reliability, two different people are determining the outcomes of the test, if both people are consistent in their scores, in their outcomes of the test, then it is reliable. In parallel-forms reliability, tests are compared to each other but they were made from the same