Design of Experiments > Convergent Validity and Discriminant Validity
You may find it helpful to read this article first: What is Construct Validity?
What are Convergent Validity and Discriminant Validity?
Convergent Validity is a sub-type of construct validity. Construct validity means that a test designed to measure a particular construct (i.e. intelligence) is actually measuring that construct. Convergent validity takes two measures that are supposed to be measuring the same construct and shows that they are related. Conversely, discriminant validity shows that two measures that are not supposed to be related are in fact, unrelated. Both types of validity are a requirement for excellent construct validity.
Let’s say you were researching depression in college students. In order to measure depression (the construct), you use two measurements: a survey and participant observation. If the scores from your two measurements are close enough (i.e. they converge), this demonstrates that they are measuring the same construct. If they don’t converge, this could indicate they are measuring different constructs (for example, anger and depression or self-worth and depression).
A specific example can be found in Chou et al. (2005), who studied the Chinese version of the Geriatric Suicide Ideation Scale. The sample consisted of 154 older adults. The summary concluded that “In terms of convergent validity, the GSIS-C correlated significantly and positively with depression (assessed by CES-D), loneliness (assessed by Revised UCLA Loneliness Scale), and hopelessness (assessed by Beck’s Hopelessness Scale.”
Convergent Validity, Discriminant Validity and Correlation
Defining constructs can be a challenge. Therefore, applying convergent validity and discriminant validity can also be a challenge. Convergent validity is usually accomplished by demonstrating a correlation between the two measures, although it’s rare that any two measures will be perfectly convergent. In the case of discriminant validity, you could show that there is no correlation at all.
Correlation is measured by a correlation coefficient, r, on a scale of -1 to 1, where r=-1 is perfect negative correlation, r=1 is perfect positive correlation, and r=0 is no correlation at all.
Convergent validity is sometimes claimed if the correlation coefficient is above .50, although it’s usually recommended at above .70.
Carlson & Herdman, 2012. Retrieved Feb 3, 2016 from: www.management.pamplin.vt.edu/directory/Articles/Carlson1.pdf.