In the academic landscape, researchers, students, and educators are constantly engaging with research and assessment methods in the work they undertake. Conducting experiments, using different kinds of modeling as well and testing theories often depend not only on how accurately they are measured and validated but also on whether they have replicability. In this article, we will try to understand what convergent validity is and how it is measured and used.
What is convergent validity?
Convergent validity assesses how strongly two or more measures or responses are related to each other while using a similar construct or concept. The American Psychological Association (APA) defines convergent validity as “the extent to which responses on a test or instrument exhibit a strong relationship with responses on conceptually similar tests or instruments.”¹ The relationship or correlation of the two measures or tests have to align in convergent validity while using the same construct or concept. Here, the term construct refers to a concept, approach, mindset, or conduct that is not discernible straight away.
Let us try to understand this using a simple example. Imagine a company developing a new employee satisfaction survey to assess how content employees are in their roles. To establish convergent validity, the company compares results from the new survey with results from a well-established measure of job satisfaction, like employee retention rates or employee engagement scores.
Suppose employees who report high satisfaction on the new survey also have higher retention rates or engagement levels (e.g., participation in team activities, volunteering for projects). In that case, it indicates that the survey has convergent validity. This is because job satisfaction should logically align with behaviors like staying with the company or being actively engaged at work. This ensures the company can confidently use the new survey in employee feedback programs.
How to measure convergent validity?
Let us now see how convergent validity is measured. Like other validity methods, convergent validity can be measured by using different statistical methods. By using correlation analysis (Pearson’s (r), one can calculate the correlation coefficient between the two measures. This is one of the most commonly used methods to measure convergent validity. As explained, to assess the convergent validity of a test, we need to show a positive correlation between measures of constructs that are common or related to one another.
Therefore, if there is a high score on one scale, then a high score should also reflect on the related scale. Here, the relationship can be quantified using a correlation coefficient (Pearson’s r) where the value ranges from -1 to 1, indicating both the strength and direction of the relationship between the variables. Researchers establish a ceiling for acceptable correlation coefficients or factor loadings to affirm convergent validity. While these vary. Generally, a number more than 0.5 is considered apt and confirms convergent validity.
One can also use factor analysis, Exploratory Factor Analysis (EFA), and Confirmatory Factor Analysis (CFA) to calculate convergent validity. While EFA can be employed to ascertain the causal factor structure of a set of measures, CFA tests if the measures load onto the theory-based factors. In CFA, average variance extracted (AVE) values aid in assessing convergent validity, where values more than 0.5 show satisfactory convergent validity. Another method we can also use is structural equation modeling (SEM) to look at the relationship between measures and the fundamental construct. As SEM blends both regression analysis and factor analysis, it is viewed as a superior technique for measuring convergent validity.
How does convergent validity differ from discriminant validity?
Convergent validity and discriminant validity are the two sub-types of construct validity in the field of social sciences. Construct validity is the degree or extent to which a test is able to measure a concept or theory that it was designed to measure.
Often, researchers evaluate both convergent and discriminant validity together to understand if a test actually measures the concept or trait it was intended to measure. As we have seen, convergent validity determines whether a test intended to measure a specific construct is correlated with other tests that evaluate the same or similar constructs. At the same time, discriminant validity assesses whether two tests that measure different constructs are expected to be unrelated and do indeed show that there is no correlation.
Hence, to prove that there is high construct validity, a test or measure should reflect high discriminant validity and equally high convergent validity. Let us look at an example to understand discriminant validity. A school administers two tests to understand their students’ social behavior. One test asks questions regarding their ease of communication with strangers, while the other test evaluates their basic mathematical skills. If the two test scores are compared, you will find that there is absolutely no correlation between them as they measure totally different constructs. The two tests are based on unique constructs totally unrelated to one another, pointing to discriminant validity.
Understanding the concept of convergent validity is essential for early-career researchers to build credibility in their work and ensure robust, reliable results. By comparing new measures with established tools or related parameters, researchers can confirm that their findings align with theoretical expectations.
References:
- https://dictionary.apa.org
R Discovery is a literature search and research reading platform that accelerates your research discovery journey by keeping you updated on the latest, most relevant scholarly content. With 250M+ research articles sourced from trusted aggregators like CrossRef, Unpaywall, PubMed, PubMed Central, Open Alex and top publishing houses like Springer Nature, JAMA, IOP, Taylor & Francis, NEJM, BMJ, Karger, SAGE, Emerald Publishing and more, R Discovery puts a world of research at your fingertips.
Try R Discovery Prime FREE for 1 week or upgrade at just US$72 a year to access premium features that let you listen to research on the go, read in your language, collaborate with peers, auto sync with reference managers, and much more. Choose a simpler, smarter way to find and read research – Download the app and start your free 7-day trial today!
