Validity refers to the extent to which a test or measurement tool accurately measures what it’s intended to measure. For instance, if a test of intelligence measures intelligence, then it should only measure intelligence and not anything else. It tells you how accurately a method measures something. This article discusses the concept of validity and its different types, along with examples.
Value of validity in research
It’s different types have significance in the process of research. The absence of valid findings leads to false conclusions and wrong decisions. It is considered essential in research on the following grounds:
- It is made sure that the findings of the research precisely represent the study concept so as to enhance the dependability of the research.
- Valid results are replicated by other researchers and are used to build upon knowledge.
- The valid results can be used successfully in real-world situations.
- Because of it, bias and errors are reduced in the process of collecting information and its analysis.
Types
It has the following four major types:
Construct validity
This type ensures that the tool of measurement represents the thing in which we are interested in measuring. It is used to establish the validity of a method as a whole.
Concept of construct
Construct indicates a concept that cannot be directly observed; rather, it is measured by other associated indicators. Intelligence, job satisfaction, depression, and memory are examples of constructs. Depression cannot be directly measured; rather, it can be measured through its indicators, such as low confidence or energy levels. So the data collection tool, such as a questionnaire, should contain items related to the indicators of depression.
Instead of consisting of simple, factual questions, the complex questions are framed to check the theoretical and observational terms. It is imperative to make sure that the phenomenon being investigated actually exists. This phenomenon should have the power to predict relationships. The researcher should demonstrate the evidence as much as possible. There are various methods for determining the test’s construct validity.
Convergent validity
As a subtype, convergent validity measures the degree of theoretical relationship between the two measures. It shows that measures of the same constructs have a high correlation. For example, if two diverse tests A and B are used for measuring the self-esteem of a person, then the performance noted on test A should be the same as the performance measured on test B, and vice versa. The strong correlation between the results of these two tests shows proof of its existence.
Concurrent validity
It assesses the extent to which the results of a test match the results of already established and accepted tests during their administration simultaneously. It determines the extent to which the new test appropriately reflects the already established test without waiting for observation of the results in the future. So in the case of validating a new test in comparison to the currently existing criteria leads to kind.
Predictive validity
It measures how accurately a test forecasts a standard or criterion in the future. It expects the individual’s performance in relation to a measured criterion later on. For example, on the basis of a new intelligence test, a prediction can be made that a scorer of age 12 will obtain a higher degree in later years.
Content validity
It calculates the extent to which the test represents all the indicators of a construct. Content validity examines the adequacy of covering all aspects of test items. To produce valid findings, it is essential that all the indicators of the subject are included in the test. In the event that some indicators are missing from the test, it is endangered, and it is expected that the research suffers from omitted variable bias.
Content validity, example 1
A content validity example can be that if a mathematics teacher prepares an algebra test, then it is essential for him or her to cover all forms of algebra taught in the class. In cases where some parts are missing, the students’ understanding of the subject may not be accurately indicated. In the same manner, irrelevant questions will also mark questions on the content validity of the test.
Content validity example 2
Another Content validity example is that of a mobile phone company that wants to check the customer satisfaction level regarding a newly launched phone model. The company will conduct a survey comprising questions depicting the features, performance, color, design, price, etc. of the new mobile model. If some questions are left out and some aspects of satisfaction are missing, then the results will be invalid.
Face validity
The face validity of a test shows its appearance or face value. It considers whether the content of the test seems to be on the surface. It is more subjective and informal. If the purpose of the test is clearly mentioned, then it has high face validity, and vice versa. test, asking the respondents to rate it. A Likert scale can be used by the respondent to assess it.
Face validity example 1
- There is extreme suitability of the test for the given purpose
- The test is suitable for the given purpose
- The test is acceptable
- The test is unacceptable
- The test is irrelevant and, therefore, unsuitable
If there is a greater level of agreement among raters, then it can be considered a strong construct. When the experts in the field are consulted for ratings, then content should be used rather than face validity.
There is a possibility that the test item may have no face validity, yet it may have a general one. This is advantageous because the respondents cannot easily manipulate the answers.
Face validity example 2
If a teacher intends to know about the mental capability of the student, he can create a survey and ask him questions regarding his time spent studying. Here, the survey represents the exploration intentions of the teacher. This item represents the high-face rationality of the survey.
Face validity example 3
The other face validity example is the measurement of the regularity of the individuals’ dietary habits. The researcher will ask the respondents about their habits for every meal of the day and snacks taken every week. The survey items will be good because they represent what the researcher wants to measure.
It is important to remember that this is the weakest kind of validity because it is a subjective measure. Yet it is considered useful during the initial stages of the development of a method.
Criterion-related validity
It gauges how well some tests can predict an outcome or how well the results of one test estimate another’s results. In other words, during its evaluation, the researcher calculates the relationship between the results of his measurement and the results of the criterion measurement. The case of high correlation indicates that a test measures what it claims to measure.
Criterion-related validity example 1
One criterion-related validity example can be the creation of a test by a university professor for measuring the English writing ability of the applicants. For assessment of the test’s ability to measure the students’ writing ability, he finds an already existing test considered valid for measuring the students English writing ability and compares the results when the same students take both tests. In cases of similar outcomes, the new test is considered to be highly valid.
Criterion-related validity example 2
Another criterion-related validity example is the creation of a test for measuring the capacity of job applicants to create ads. The manager discovers another test that is considered a standard for measuring the applicant’s capability to create ads.
The manager will use the standard test as a judgment criterion. Then the results of both tests are compared to assess their capacity. In cases of similar results, the test is considered to be highly valid with reference to criterion.