Table of Contents
For proper and objective research to be achieved there is the need to have valid and reliable tools or instruments. The questions used in collecting data from the field should be valid and reliable.
Statistically, validity is the level/extent on which a question/concept, measurement, or conclusion is logically based and corresponds directly and precisely to the situation in question. Validity is the degree to which a question or rather a test corresponds to what it is supposed to measure. For a test/question it is very important to be valid so that the results are scrupulously evaluated, utilized and interpreted. Accurate collection of data from the research field directly corresponds to good data analysis and interpretation. This paper will look at two types of validity, concurrent validity and face validity. Concurrent Validity manifests when the criterion measures are taken or obtained at the same time with the test scores (Morse et al, 2002). It shows the level to which the test results measure exactly an individual’s present situation/state with regard to the available criterion. For example, a test that determines the level of depression is likely to be of concurrent validity if it measured the actual level of depression suffered by the by the test taker. Face validity estimates the ability of a test to measure a particular criterion. It is the accurateness and meaningfulness of conclusions obtained from the research results. Golafshani (2003) defines face validity as the degree to which results obtained from the analysis of the data represent the phenomenon under the study.
Reliability can be defined as the relevance of the results obtained after taking a test for the first time, to be the same when administered at different times and by different persons. In a nutshell, a test is said to be reliable if same results are obtained repeatedly. For example, if a test is meant to measure a trait depression in psychology, then each time the test is administered to a subject, the results should be almost the same. Reliability test can be done in a number of ways which include: parallel forms reliability, internal consistency reliability and test-retest reliability. This paper will use the test - retest reliability. Test-retest reliability means taking the test two times at two different points in time. This method of reliability test is used to test the relevance of a test over time (Golafshani, 2003). However, this method of reliability test assumes that there is no change that occurs in the quality or constructs being tested. Generally, reliability is higher if little time is allowed between the two tests.
- Reliability test is then done by me using the test- retest method to establish if the test is reliable. If the scores obtained during the test every time it is taken differ a lot with each other, then the test is not good to predict anything. However, if a test is reliable, that does not mean that it is valid.
- The scrutinized questions will be put to make judgments of their face validity. They may be presented to another specialist to scrutinize them and also make judgment of their face appearance. In other words, this is a check of whether the questions are related to the issues to be researched for the case, The risk of teenagers driving while intoxicated. If the two or more specialists make the same judgment or almost the same judgment, then the questions are valid.
- The questions are further put under concurrent validity test. Coefficients of the questions will be calculated and related to the criterion and thus validity is established. If the coefficient found is 1.0, then the questions are valid.
Relationship between the results of reliability and validity obtained in the consultancy
Limited time Offer
If data collected are valid, they must be reliable. Reliability is a good measure, but not dependable and sufficient condition for validity. That is, if a test is valid, it is reliable but not all reliable tests are valid.
From the results of the above tests of the question of the research, M.A.D.D can go on to carry out the research not depending on the results validity and reliability obtained from the consultancy.