Establishing Reliability and Validity Essay

869 Words4 Pages
Establishing Reliability and Validity In conducting a research or survey, the quality of the data collected in the research is of utmost importance. One’s assessment may be reliable and not valid and thus this is why it is important that when designing a survey, one should also come up with the methods of testing the reliability and validity of the assessment tools. For MADD (Mothers Against Drunk Driving) to conduct a survey, the questions they propose to use must pass the validity and reliability test for one to conclude that the survey is reliable and valid. This survey will try to find out the risk factors that contribute to drunken driving by teenagers or young adults. Reliability can be defined as the statistical measurement of…show more content…
On the other hand, the types of validity include content validity, criterion validity and construct validity (Litwin, 1995). The assessment of these forms of reliability and validity determines the quality of the data that our tools will collect and hence affects how reliable and valid the research will be. When using multiple indicators, the test-retest is the most common and easiest. This is usually done by administering survey questions to the same respondents at different times so as to see how consistent their responses are (Litwin, 1995, p. 8). This process measures how reproducible the results are. When the two sets of responses from the same respondent are compared, their correlation is referred to as intraobserver reliability. This measures the stability of the responses from the same respondent as a form of the test-retest reliability. The alternate-form or alternative method is almost similar to the test-retest method but differs on the second testing, where instead of giving the same test an alternative form of the test is given to the same respondents (Carmines & Zeller, 1979, p. 40). However, the two tests should be equivalent in that they should be designed to measure the same thing. The correlation between the results of the two forms is the interobserver test, which gives an estimate of the reliability. The split-halves test involves splitting the survey sample
Open Document