If a test serves what it is intended to do, we say that it is valid. For example, a test Intelligence it is valid if it measures intelligence. Reliability is the degree to which an assessment tool produces stable and consistent results. The creliability of the tests it is a measure of reliability obtained by administering the same test twice over a period of time to a group of individuals.
Tests are valid insofar as they measure what they are supposed to measure. We can distinguish between face validity and proven validity:
- Apparent validity: refers to the degree to which the test appears to measure what it wants to measure.
- Proven validity: is the one that has been verified in some empirical or experimental way. The proven validity is empirical or theoretical.
- Empirical validity: it is empirical when it has been verified in a practical way. For example, a driver selection test is valid if it distinguishes good from bad drivers, despite not knowing exactly why it succeeds.
-
Theoretical validity: reflects the degree to which it has been proven with scientific arguments that the test measures the property or trait that it tries to measure.
For example, an intelligence test will have theoretical validity when there is a proof of it that is supported by a logical and experimental verification.
- Face validity: A test has face validity if it appears to measure what it is trying to measure. It is not a sufficient validity, although it is often necessary. Many times, it is convenient that the test seems valid to the subjects who have to answer it. At other times, it is essential that the test does not appear to measure what it measures. This is the first requirement of effective validity. To achieve this, it is not enough for the test to appear valid, it must be verified that it is.
- The verified validity: The effective validity is the verified validity. There are many ways to check validity, some are empirical and others are logical-experimental.
- The empirical or criterion validity: A test has empirical validity if it has been proven that it serves some practical purpose. For example, a test is valid for selecting drivers if it distinguishes between different degrees of driving skills.
The empirical validity it is also called criterion validity, and to check it, an external validity criterion must be defined. In the example discussed, the external criterion is driving skills. The test validity coefficient is the correlation between the test and the criterion. If the best in the test are those who drive the best, the test will have a high coefficient of validity with respect to that criterion, and we will have an empirically valid test to select drivers.
This validity is scientifically blind, because we do not know what the validity of the test is based on. The test is valid for selecting drivers, we do not know why. The empirical validity can be classified as:
- Prospective or predictive or prognostic validity: It is the one that is verified by means of the correlation between the test and a later measured criterion. For example, if we have many aspiring pilots and we want to select those who are more likely to become, after studies and practices, good pilots. We can apply certain tests that appreciate the requirements of being a good pilot, and admit all applicants. After the studies and the opportune practices, we will measure the subjects in the criteria that guarantee to be a good pilot. If the tests answered by the subjects at the time of admission have a high correlation with proficiency as pilot, measured after learning, perhaps years later, it can be affirmed that these tests have high validity prospective. This validity has been called prospective, predictive and prognostic validity, since it allows to conveniently forecast or predict the results of the subjects in the criterion, serves to select the applicants who are most likely to be good pilots.
- Inspective or concomitant or concurrent validity: It is generally too expensive to test predictive validity; it is usually outside the practical possibilities of the psychologist. For this reason, the method of inspection, concomitant or concurrent validity is frequently used. It is the same as the previous procedure but the tests and the criterion are measured in the same period concomitantly or concurrently. For example, a timely sample of pilots is chosen, the appropriate tests are applied to them, and they are measured on the criterion of expertise. The coefficient of concomitant validity will be the correlation between the tests and the criterion. It informs us to what extent the good and the bad in the test are, at this moment, good and bad in the criteria. It does not guarantee us that the applicants with the best results in the tests afterwards will be the best drivers.
- Retrospective validity: It is the correlation between the tests, applied at a certain moment, and a criterion that was previously measured, perhaps years before. Its purpose is to predict a past criterion, to find out from the present the factors and conditions that in the past influenced the subjects, and that explain some of the current characteristics of the subjects.
- Theoretical validity: A test has theoretical validity if you measure what you intend to measure.
The amount in which it measures it is indicated by logical and experimental arguments that are equivalent to its relationship with an internal criterion. There are various types of theoretical validity, and they are as follows
- Content or sample validity: Mainly, it refers to tests of performance, knowledge or professional competence. For example, a spelling test is valid if its content is appropriate. To verify this validity, the field of spelling will be examined and it will be verified that the items that make up the test are an impartial and sufficient sample of this field. This type of validity cannot be specified in any type of correlation. It expresses the relationship between the test and the internal criteria formed by the subject to which the test refers. The ratio will be higher as all the matter is better represented by the test. This validity is achieved to the extent that there are guarantees that the various aspects of a certain field of knowledge or skills that the test attempts to measure have been clearly defined. Criticizing the sample validity of a test implies showing that it does not adequately represent the field to which it refers.
- Construct or conceptual validity: It consists of verifying, according to the methodology of scientific research, that the test measures the variable to which it refers. The aim is to scientifically guarantee that the variable that the test tries to measure is an acceptable variable, the concept of which presents sufficient logical consistency within the theoretical system of psychology and is supported by sufficient experimental evidence that verify. For example, before offering a valid test to measure intelligence, it will be necessary to clarify what that intelligence is that the test tries to measure.
- Factor validity: It occurs when different tests are correlated with a certain trait and then subjected to a factorial relationship.
- Congruent validity: If we measure different variables with tests of the same type, the tests that refer to a variable must have higher correlations with each other than they have with the tests of other variables.
- Discriminatory validity: If we measure different variables with tests of the same type, and the same variable with tests of different types, these tests of different types do reference to the same variable must have higher correlations with each other than tests of the same type that refer to different variables.
- Structural validity: It is the degree to which the test measures an elaboration or a theoretical trait. Any data that gives some information about the nature of the trait we are considering and of The conditions that affect its development and manifestations are useful for this type of validity.
This article is merely informative, in Psychology-Online we do not have the power to make a diagnosis or recommend a treatment. We invite you to go to a psychologist to treat your particular case.
